AN ABSTRACT OF THE THESIS OF

GERARD ADRIAN BERTR.AND for the DOCTOR OF HILOSOPHY (Name ofstudent) (Degree) inOceanography presented on /P) j9'2O (Major) (Date) Title: A COMPARATIVE STUDY OF THE INFAUNA OF THE

CENTRAL OREGON CONTINENTAL SHELF Redacted for privacy Abstract approved: ndrew G. C)arey, Jr

One hundred and sixty 0, 1 meter2 Smith-McIntyre grab sam- pies were taken on the Oregon Continental Shelf at eight seasonal stations between 75 and 450 meters depth.The five replicate grab samples per season per station were analyzed for macrofauna (>1.0 mm). Particular attention was paid to shelled , Cumacea, and Ophiuroidea. The samples were analyzed for total , number of specimens, wet weight, and ash free dry weight0 Seventeen environmental parameters were measured at each station at each season. Results showed no seasonal variation in either the infaunal composition in total species, numbers, or biomass, or in the sedi- ment environmental parameters. The average values for all stations over the year-long study were 597 individuals permeter2,36.5 grams wet weight permeter2,and 2. 57 gramssh free dry weight per-meter2.These values are lower than those reported for Southern California and for the-Northeast coast of the United States. Four species groups which were the dominant fauna in beach sand, silty sand, sandy silt, and glauconite sand were extracted by

R. mode factor analysis. - The Q mode factor analysis showed two distinct sand communities in glauconite sand and beach sand.The silty sand stations had high loading to two factors while the sandy silt stations had high loading to no factors. A fourth community of organisms was described for the Oregon continental shelf in addition to the three previously described off Washington. The results of regression analysis on nine major environmental factors showed no meaningful correlation that accounted for more than 39 percent of the total variation. Only seven of the 21 most abundant molluscan species showed meaningful significant correlation to one or more-of the environmental variables.This may be an indication of biotic interdependence. A Comparative Study of the Infauna of the Central Or egon Continental Shelf

by Gerard Adrian Bertrand

A THESIS submitted to Oregon State University

in partial fulfillment of the requirements for the degree of Doctor of Philosophy June 1971 APPROVED: Redacted for privacy

Assistant Prossor of O'ceanography in charge of major Redacted for privacy

Chirman of Departnent of Oceanography

Redacted for privacy

Dean of Graduate School

Date thesis is presented !724U /J /97) Typed by Opal Grossnicklaus for Gerard Adrian Bertrand ACKNOWLEDGEMENTS

A large scale synecological study such as this could not be the work of one man. Geologists in the Department of Oceanography at Oregon State did the sediment particle size analysis and chemists did the water nutrient analysis.Aid in identification was given by Mr,, Michael Kyte for the ophiuroidea, Dr. Myra Keen the bivalve mol- luscs, and Dr. Ulf Lie for the cumaceans. Dr. Scott Overton provided the extremely helpful AIDN com- puter program for analyzing the data and through discussion gave assistance in interpreting the results.Mrs. Jonna Zipperer did the legwork to modify the AIDN program for my data; she also ran the analysis. Dr. Wayne Courtney and Dr. Jack Corliss counseled me on the use and abuse of factor analysis.Dr. 1511 Lie provided unpub- lished data on the infaunal communities on the Washington continental shelf.Numerous members of the Ralph York Society, and in par- ticularMr. Frederick Gunther, aided in collecting samples on the R. V. Cayuse. I thank Dr. Charles Miller and the members of my committee, Dr. Joel Hedgpeth, Dr. William Chilcote, Dr. John Wiens, and Dr. David Bostwick for critically reading the manuscript and pro- viding valuable suggestions. I am especially grateful to my major professor, Dr. Andrew G. Carey, Jr., for assistance in every phase of the study and for graciously accepting my eccentric work schedule. Without help and encouragement from my wife, Faith, the work would have been neither started nor completed. Support was provided by Dr. Carey and by the Sea Grant Pro- gram, Grant GH97. TABLE OF CONTENTS

L INTRODUCTION

MATERIALS AND METHODS 5

The Study Area 5 Station Selection 7 Choice of Sampler and Sieve Size 9 Number of Samples 14 Environmental Sampling 16 Laboratory Processing 17 Wet Weights and Ash Free Dry Weights. 19 Laboratory Environmental Sample Processing 20 Particle Size Analysis 20 Water and Sediment Analysis 22 Data Analysis 23 Some Computational Definitions and Methods 24 Factor Analysis 27

RESULTS 29

The Benthic Environment 29 Station Summary 35 Station 2 35 Station 6 35 Station 7 36 Station 8 36 Station 10 36 Station 15 37 Station22 37 Station 23 37 Environmental Seasonality 38 Sampling Results 39 Species Distribution and Abundance 43 In-Station Variation 53 Station-to-Station Variation 61 Season-to-Season Variation 68 Standing Stock 70 Factor Analysis 78 R Mode 78 QMode 83 Regression Analysis 89

DISCUSSION 94

CONCLUSIONS 105

BIBLIOGRAPHY 107

APPENDICES 114 LIST OF TABLES Table Page The nine seasonal Sea Grant stations and cruise dates. 10 Preserved wet weight to ash free dry weight conversion factors. 21 The 17 sampled environmental parameters by station. 30 Percent of the total number of species and total number of individuals from the added species. 42

Number of individuals per meter2 by major taxa. 47 Species code list for all molluscs, ophiuroids and cumaceans. 48

Within station comparisons of DIF, SIMI, andSd2. 59 Two measures of weighted mean niche breadth (Levins, 1968) within station by season. 60

Similarity index (SIMI) between stations. 62

Difference index (DIF) between sta.tions. 63

11, Two measures of weighted mean niche breadth (Levins, 1968) summed for all seasons within a station. 67 Difference index (DIF) and similarity index (SIMI) between seasons summed for all stations. 69

Seasonal variations inSd2,He, B2, B5 summed for all stations, 69 Ash free dry weight in grams permeter2of the major taxa by station, 76 Table Page

15. The 15 factors extracted by R mode factor analysis: eigenvalue, percent total variation explained, and cumulation percent total variation by factor. 80 The four species groups extracted by R mode factor analysis. 81 Results of Q mode factor analysis with four factors rotated, 85 Regression analysis for the 21 most abundant mol- luscs on nine environmental parameters. 91 LIST OF FIGURES Figure Page Distribution of sediment types and bathymetry. One mile equals 185 kilometers,1 fathom equals 1 . 83 meters (Byrne and Panshin, 1968). 6 The 16 initial Sea Grant stations sampled in July, 1968. One nautical mile equals 1.85 kilometers; 1 fathom equals 1.83 meters. 8 The nine Sea Grant stations sampled seasonally. One nautical mile equals 1,85 kilometers; 1 fathom equals 1.83 meters. 11

Sediment particle size distribtuion by station. 1 2 The correlation coefficient matrix for all stations at all seasons. 33 Cumulative number of species of molluscs, ophiuroids and cumaceans with increased sampling (10 0. 1 m2 samples). 40 Total number of identified species and individuals of molluscs, ophiuroids and cumaceans per two meters2. 44 Distribution of the total number of individuals by major taxa. 46 Distribution of major molluscan species at stations 8, 10, 2 and 6. 51 Distribution of major molluscan species at stations 22, 23, 7 and 15. 52 Within station variation of similarity index (SIMI) during the sampling period. 54 Within station variation of difference index (DIF) during the sampling period. 56 Within station variation of Sd2 during the sampling period. 58 Figure Page and 1 4. Comparison of the Shannon-Weirer function, -e Simpsont s diversity index, Sd2 for all stations. 65 A two dimensional representation of the relationships between stations based on the greatest value of SIMI between stations (upper) and the lowest value of DIF (bottom) 66 Seasonal variation in the number of bivalves for all stations. 71 Wet weight in grams permeter2of the edible fraction of the biomass within all stations. 72 Ash free dry weight in grams permeter2 of the edible fraction of the biomass within all stations. 74 Total ash free dry weight in grams permeter2of the major taxa by station. 75 Seasonal variation in grams permeter2 of thewet weight of the edible fraction of the biomass. 77 Seasonal variation ingrams permeter2of the ash free dry weight of the edible fraction of the biomass. 79 Interrelationships of the eight seasonal stations. One nautical mile equals 1. 85 kilometers; 1 fathom equals 1.83 meters. 90 A COMPARATIVE STUDY OF THE 1NFAUNA OF THE CENTRAL OREGON CONTINENTAL SHELF

I.INTRODUCTION

Although the infauna of the southern California and nea.rshore Washington continental shelf has been studied in some detail, there has been little done in Oregon on the distribution, abundance and sea- sonal variation of the infaunal shelf communities. Knowledge of the composition of these communities and their seasonal changes could form the base for future investigations of benthic production.Studies on selective feeding by demersal fish or of changes in feeding habits with seasons, sex, or size must also have as their base a thorough knowledge of the major food source, the infauria.For these reasons a study of the infauna off the central Oregon shelf was undertaken in July 1968.The study had as its objectives:the determination of 1) the composition of infaunal communities in a wide variety of shelf sediment types, in numbers of individuals and species, 2) the stand- ing stock of the infaunal components, 3) the extent of seasonal varia- tion in species composition, numbers, and standing stock, 4) there- lationship of the infaunal communities to each other and to various sediment types and depths, and 5) the relationship of major species in numbers and frequency of occurrence to a wide variety of environ- mental parameters. 2 Quantitative subtidal benthic investigations were first initiated by Petersen and Jensen (1911) using a small half-moon shaped grab that sampled an area of 0. 1m2.Based on his studies in Denmark Petersen (1913, 1918) advanced the concept of benthic communities as statistical units.This concept became widely known and accepted by terrestrial ecologists.The advantage of the Petersen grab over previous methods used to investigate the benthos was that it enabled the investigator to enumerate such faunistic parameters as number of species, number of individuals, and standing stock per unit area, The utility of the Petersen method of mapping and enumerating the benthic communities led to its use in a large number and wide variety of other areas.Packard (1918) presented work on the mol- luscs obtained with an orange peel grab by the U, S. Albatross in 1912 and 1913 in San Francisco BaysDavis (1923, 1925) conducted studies on the bottom fauna on Dogger Bank in the North Sea.And Ford (1923) investigated benthic communities in the English Channel. Zenkevitch etal. (1928) attempted to estimate the productivity of the benthos in the Kara and Barents Seas.The relationship of the benthos to the demersal fish population was investigated by Steven (1930) using a quantitative grab similar to that used by Petersen Because the distribution of the benthos was patchy, Steven took five grab samples at each station and combined them.Working with mol- luscs off the coast of Greenland, Thorson (1933, 1934) confirmed the idea of benthic communities based on dominant members. Stephen (1933, 1934) and Lindroth (1935) first rejected the idea of benthic com- munities because of the difficulty of delineating the boundaries between them. The first attempt to delineate benthic communities in the United States was made by Shelford and Towler (1925) using Petersen grab samples obtained in Puget Sound.Shelford etal. (1935) continued this survey work in Puget Sound covering a large area butwith a smalinum- ber of samples. Hartman(1955) used an orange peel bucket as a sampler to survey the bottom fauna of the San Pedro Basin off South- ern California.The extensive study by Barnard and 0. Jones (1959) for the State Water Quality Control Board of California delineated the species composition and distribution by sediment types and depth on a large area of the Southern California shelf.Meredith Jones (1961) working in a small area off Port Richmond, San Francisco Bay, California occupied four stations every five to eight weeks for a period of 14 months. Although the number of stations M. Jones used was small, the number of samples was very large; he was able to estimate the distributional patterns of the species studied. Lie (1968) conducted an extensive investigation of the benthic infauna of Puget Sound using the Van Veen grab.This quantitative study yielded a great deal of information about sampling efficiency, biomass distribution, species distribution, and much useful 4 information on conversion factors from wet weight to ash free dry weight.The work is distinguished by the large number of samples and completeness of the data.Lie and Kelly (1970) extended this work to the mouthoi Puget Sound and to the open Washington coast. Using factor analysis on the 35 most abundant species they were able to distinguish three distinct communities lying parallel to the coastline in sand, silty sand, and silt substrates.The only work on the Oregon continental shelf is that of Carey (in press) on the inter- relationships of eight stations off Newport, Oregon from 25 to 200 meters.In this work the similarity between the three stations is computed based on polychaete species distribution, numbers, and biomas s. II.MATERIALS AND METHODS

The Study Area

The area chosen for investigation lies between 75 and 450 me- ters on the Oregon continental shelf between the Umpqua River and the Yaquina River (Figure 1 )It is a region of diverse sediments and is well-suited to a study of -sediment relationships. The most characteristic feature of this area of the continental shelf is Heceta Bank.This bank is 33 kilometers long and up to 15 kilometers wide, and it dominates the outer portion of the shelf(Maloney, 1965).The eastern portion of the shelf inside the bank is very gently sloping, smooth and sediment covered.The primary sediment of the inner portion of the shelf is clean, well-sorted detrital sand that occurs to a depth of 100 meters. These sands grade into silty sands and silts withglauconitic sands predominating on the topographic lows on the outer shelf.The two primary sources for sediments on the Oregon shelf are rivers and the erosion of coastal terrace deposits (Runge,

1965).Rivers contribute the finer grain sediments while the terrace deposits contribute the coarse grain materials, The complexity of the study area is increased by the deposit of fine grain sediments near shore from the river mouths.Unlike the coast of Washington the study area has considerable heterogeneity of sediment distribution. 6 :

O HEAD

::::, T/LL4AIOOK BAY

YAQU/NA YAQUINA SAY BAY

UMPQIJA RIVER

ROCK Coos Coos SAND BAY BAY El MUDDY SAND MUD

0 0 20 0 10 20 NUT. MI. NAUT. MI.

CONTOUR INTERVALIOFMS. CAPE BL ANCO

Figure 1.Distribution of sediment types and bathymetry.One mile equals 1. 85 kilometers,1 fathom equals 1. 83 meters (Byrne and Panshin, 1968). 7 The entire shelf area is characterized by high production from up- welling during the summer months and by high river run-off and deposition during the winter months.This run-off reaches a peak during the spring when melting snows make their contribution,

Station Selection

During July1968the initial Sea Grant cruise sampled 16 sta- tions (Figure 2) to determine sediment homogeneity within a single station.Only those stations with uniform sediment distribution were chosen as seasonal stations.Upon arrival at station, a buoy with a radar reflector was moored to ensure accurate navigation while on station.Station position was determined by Loran C and a Precision Depth Recorder.Position fixes were taken between each grab sample and the ship's drift corrected to maintain position.Samples were taken at the center of the station and at the apices of an equilateral triangle one nautical mile (1. 85km) to a side with the moored buoy as its center. One or two Smith-McIntyre grabs (Smith and McIntyre,

1954)were taken at each apex and one Smith-McIntyre grab from the center of the station.In addition an anchor-box dredge (Carey and

Hancock,1965)was taken at the station's center.Quantitative beam trawis (Carey, in press) were taken between apices of the triangle. Homogeneity of the sediments at each station was determined by particle size analysis of Phleger cores (Fowler and Kuim,1966) )::VAQU/NA 21 :.aAY h 44°30 1 4 26

'i 22 :o_ '? - STATUTE MILES

6. 44°OO' 24 8. ..I0 25 2j UMP QUA RIVER

125°CO' 124°30' I 24°OO Figure 2.The 16 initial Sea Grant stations sampled in July, 1968. One nautical mile equals 1 .85 kilometers; 1 fathoms equals 1.83 meters. 9 taken in conjunctionwith the Smith-McIntyre grabs. Of the original 16 stations sampled nine were selected for seasonal study. One of the seasonal stations, station 1off the Yaquina River, was abandoned because of poor weather on two occasions and was not included in the present investigation.The remaining eight stations and cruise data are given in Table 1.The distribution of these stations on the Oregon continental shelf is shown in Figure 3.These stations, as originally conceived, were in pairs to sample the same sediment type at different depths and compare the faunas.Stations 2 and 7 were one pair, 6 and 8 another, and 10 and 15 a third pair.Stations 22 and 1 would also have been a pair if 1 had not been abandoned.Station 23 was intermediate between stations 6 and 8, and 10 and 15 in sediment type.Particle size distribution within a station can be seen in Fig- ure 4.

Choice of Sampler and Sieve Size

The sampler chosen for this study was an 0. 1 meter2 Smith- Mcintyre grab.Smith and McIntyre (1954) and Wigley (1967) found this grab to be more efficient than the similar Van Veen grab (Tham- drup, 1938).Sampling for the entire investigation was conducted from the research vessel Cayuse. Because the Cayuse is an 80 foot vessel the sampling was very dependent on weather conditions The Smith-McIntyre grab was chosen as it is most reliable of the 10

Table 1.The nine seasonal Sea Grant stations and cruise dates.

Stations Latitude Longitude Depth (meters)

SG 2 44°39.O'N 124°35.9'W 200

SG 6 44°01.O'N 124°35.7'W 150

SG 7 43°57.4'N - 124°18.O'W 100

SC 8 44°55.8'N 124°35.7'W 200

SG 10 43°48.5'N 124°51.0W 450

SG 15 44°O9.0N 124°25.OW 100

SC 22 44°14.3'N 124°16.8'W 7$

SG 23 43°48. S'N 124°17. 8'W 100

Cruise Departure Date Number (1 eek duration)

C6810C October 22, 1968

C6901C January 24, 1969

C6904C April 21, 1969

C6907C July 23, 1969 11 SEA GRANT STATIONS

I Ij 125000s 124°30' Figure 3.The nine Sea Grant stations sampledseasonally.One nautical mile equals 1. 85 kilometers; 1 fathom equals 1 .83 meters. 4Cl)z0 1JI I 00000000000 N-CO IO 'o) N- CO 13 quantitative grabs in marginal weather conditions (Smith and McIntyre, 1954; Lie, 1968).Except at station 10, the deep sta- tion at 450 meters, pretripping of the sampler in the water column caused by shiptsroll was minimal. The bite taken by the Smith-McIntyre grab from the bottom was compared to those of the Van Veen and the Petersen grabs by Gallardo (1965).The Petersen and Van Veen grabs bite deeper along the edges; the Smith-McIntyre bites deepest at the center. Further, the Smith-McIntyre was most affected by substrate, tak- ing a slightly semi-circular bite in mud and a rectangular bite in sand.It is, however, more effective on hard bottoms than either of the other grabs. The grab is also most consistent in getting a constant volume from a homogeneous substrate. During the present study the grab sampler was lowered at a constant two to three meters per second until it reached a point ten meters above the bottom,The speed was then increased to five meW- ters per second.This increase in speed allowed the grab to hit the bottom solidly and minimized the effect of the ship's roll.Upon re- trieval the grab sample was placed in a hopper similar to that de- scribed by Holme (1959) and modified by Carey (Carey and Paul, 1968) for work off the Oregon coast.The grab was opened and the sample washed onto a screen with a 0. 42 mm aperture with a fine stream of sea water from a hose.Reish (1959) showed that this 14 screen size would be effective in retaining 90 percent of the infaunal specimens, species, and biomass collected.It did, however, allow most of the fine silts and clays topass through the sieve thus reduc- ing the volume to be preserved and sorted.The remainder of the sample was preserved in ten percent formaldehyde bufferedwith Na2B4O7. The sampleswere then brought to the laboratory for proc e s sing.

Number of Samples

There has been a great deal of discussion in the literature about what constitutes an adequate sample in benthicecology.Thorson (1957) suggested a "standard unit" of 0.1 m2 for 0 to 200 meters and 0.2m2for 200 to 2,000 meters.This standard unit has its draw backs, depending on the type of study underway and the goals of the investigation,If only the most abundant species with the highest fre- quency of occurrence are required to characterize the fauna, 0. 1 may be an adequate sample.If, however, the goal of the investiga- tion is the understanding of the interaction of the species comprising the fauna then a far greater number of samples is needed,The num- ber of samples necessary isa function of the uniformity of the envi- ronment, the species distribution, and the object of the study. Holmes (1953), N. Jones(1956)and Longhurst (1958) have shown the relationship between the number of samples and the cumulative 15 number of species for various types of bottom samplers.Longhurst (1958, 1959) in particular has shown the relationship between number of species and number of individuals for the Smith-McIntyre grab at two shallow subtidal stations off the Sierra Leone River on the West African shelf.His data show that five replicate 0. 1in2Smith- McIntyre grabs sample greater than 50 percent of the total number of species sampled by 20 0. 1m2replicate samples. Over 50 per- cent of the individuals at each of his stations was attributable to less than ten percent of the number of species.The data suggest that the dominant species account for a lesser percent of the total number of individuals with increasing depth. It was suggested by Longhurst (1959, 1964) that a minimum area of 0, 5m2sampled by five replicate 0. 1m2grab samples was necessary to characterize the fauna.Lie (1968) tested this hypothe- sis in Puget Sound and found that three to four 0. 1m2samples were sufficient for collecting the species thatmade up 95 percent of the total number of specimens.He also found that 75 to 85 percent of the species collected by ten 0. 1 m grab replicate samples were also collected by five 0. 1m2replicate samples. For this reason, five replicate 0, 1m2grab samples were chosen as the standard for each seasons sampling at each of the eight stations. A sixth grab was taken to insure that the minimum of five was maintained in the event a grab sample was lost, mislabeled, not preserved properly, or that 16 an insufficient volume of sediment was sampled. The amount of sediment in each sample was measured with a dipstick when the top plates of the grab were opened.Those with in sufficient volume to assure a rectangular cut into the upper 15 centi- meters of bottom were not kept.Benthic investigators have, in the past, measured the volume of each sample but used it only to explain difficulties with the sample.Since most organisms live in the upper few centimeters it is believed that a15 cm depth is sufficient to cap- ture the majority of specimens, species, and biomass (Lie, 1968). In the laboratory the field data were examined without looking at the samples and the grab that had the least volume was put aside. Occasionally a grab was imperfectly preserved and this was exchanged for the one put aside.Variations in volume were minimized by sum- ming the fauna in the five grabs at each season for analysis. A close check was made on position and it was corrected and maintained throughout the sampling period. Grabs were made consecu- tively with a maximum of one half hour between samples.

Environmental Samplin

With each set of six Smith-McIntyre grabs a number of environ- mental parameters was also sampled. A modified Smith-McIntyre grab (Carey and Paul, 1968) was used for simultaneous collection of sediment and bottom water.This modification consisted of a '.7 bottom-triggered Fjarlie bottle(Fjarlie, 1953) mounted on the side of the grab.The reversing thermometer rack on the Fjarlie bottle trips automatically when the water sample is taken.The grab was lowered to five to ten meters off the bottom and allowed to equilibrate for tenminutes before the grab was lowered and the tripping occurred. Oxygen and temperature were determined on shipboard.Salinity was determined with a Hytech Model 261 inductive salinometer in the lab- oratory. A nutrient sample was drawn into a polyethylene bottle and deep-frozen immediately.The nutrient samples were analyzed for all silicates, nitrates, and phosphates on a Technicon Model 1 Autoanalyz- er after return to port. A Phieger multiple corer with five core bar- rels was taken at each station. One core was used for particle size analysis and another was frozen for sediment nutrient studies,If the sampler failed to secure a core for the sediment parameters, they were taken from an additional Smith-McIntyre grab.

Laboratory Processing

In the laboratory the faunal samples were washed on stainless steel screens of 0. 1 mm and 0, 42 mm mesh. The macrofauna for this study was retained on the first screen and about half of the meio- fauna on the second, smaller screen.The material from the 10 mm screen was then placed in 70 percent isopropyl alcohol for later sort- ing. A screen size of 1, 0 mm was chosen because by convention this 18 is the separation between the macrofauna and the smallermelofauna, Reish (1959) demonstrated that this screen size captured over 90 percent of the total biomass and almost 90 percent of the total num- ber of species.Because of this dominance in biomass and number of species the macrofauna is thought to be indicative of the infaunal

community. The samples were sorted by placing them in a large rectangu- lar Pyrex dish on white paper and picking the out with forceps and with the aid of a 3x magnifying lamp. The procedure was then repeated.The species were first sorted into major groups; however, a number of groups that were not part of theinfauna were discarded. Large calanoid copepods and euphausiids were contaminants in the sea water used to wash the grabs on board ship.Hard bottom forms such as Hydroida, hard bottom Anthozoa and Brachiopoda were re- moved from the sample. Anomuran crabs, large asteroids, and sea pens of the genus Balticina were also removedfrom the grab samples.These groups, although soft bottom inhabitants, are epi- fauna and are more adequately sampled by trawis than by grabs. Because the infauna of the Oregon coast is poorly known, the taxonomic problems involved are large.Identification of all species in the grab samples was impractical because of time limitations; consequently, three groups, molluscs, cumaceans, and ophiuroids were chosen for detailed work. The remainder of thefauna was 19 counted and weighed but not identified to species.Thorson (1951, 1957) demonstrated the importance of molluscs in characterizing benthic communities. Molluscs form a conspicuous and abundant segment of the fauna on the Oregon continental shelf; they are thought to be indicative of the fauna as a whole.All individuals of all taxa in all grabs were counted with the exception of the poiychaetes. Polychaetes were counted from two grabs from each station.Only anterior ends were counted as individuals unless the size or identity made the individual conspicuous from the rest in which case a pos- terior end was counted. Only apparently live-caught individuals were counted as part of the fauna.

Wet Weights and Ash Free Dry Weights

Wet weights were taken for each major group within each grab sample except for the molluscs in which the three most abundant gastropod and the three most abundant bivalve species in each sta- tion were weighed separately.The samples had been preserved in 70 percent isopropyl alcohol for 10 to 22 months before weighing. Each subsample was blotted to remove excess surface moisture; a ten minute interval before weighing allowed all surface moisture to evaporate and for the asymptote to be approached in weight loss from evaporation. An H5 Type Mettler analytical balance was used for weighing,All molluscs were weighed in their shells; polychaetes 20 were removed from their tubes.Because of their considerable taxo nomic value and possible future work that can be done on the samples, the specimens were not burned to obtain ash free dry weights.Con- version factors from wet weight to ash free dry weight were used in- stead (Table 2).The conversion factors of Lie (1968), Stander (1970), and Carey (1970) are quite consistent with those reviewed by Thor son (1957). Wherever possible conversion factors for the same species or members of the same genus were used. When this was not pos- sible, conversion factors for the general group were used.

Laboratory Environmental Sample Processing

Particle Size Analysis

Sediment particle size was measured by the standard proced- ures of Krumbein and Pettijohn (1938).The fine fractions were an- alyzed by the pipette method while the coarse sand fractions were analyzed by the settling tube method according to Emery (1938). Cumulative percent of the weight of the size fraction was determined for 20 phi sizes ranging from 0 to 7,966.In addition, the Inman, Trask, and Folk and Ward parameters of median grain size, mean grain size, sorting coefficient, skewness, and kurtosis were calcu- lated.Of these parameters,. percent sand percent silt, percent clay, Inman median grain size, Folk and Ward mean grain size, and 21

Table 2.Preserved wet weight to ash free dry weight conversion factors. (Lie. 1968) (Stander, 1969) (Carey, 1970 Anthozoa 0. 113

Turbellaria 0. 133

Nemertea 0. 133

Sipunculoidea 0. 130

Eqhiupida 0. 130

Polychaeta 0. 133

Aplacophora 0. 130

Gastropoda 0.083

Pelecypoda 0.055

Scaphopoda 0.060

Crustacea 0. 150

Brisaster latifrons 0.044

Allocentrotus fragilis 0,037

Ophiuroidea 0. 122

Holothuroidea 0. 076 22 Folk and Ward sorting coefficient (Shepard, 1963) were selected for

correlation with the fauna. A total of 32 seasonal samples from the eight Sea Grant stations was analyzed and an additional 56 samples from the initial 16 stations were analyzed to determine sediment homogeneity within a station.

Water and Sediment Analysis

Replicate analyses were run on each of the frozen nutrient sam pies obtained from the Fjarlie bottle attached to the Smith.-Mclntyre grab.The frozen sediment sample from the Phleger corer was an- alyzed for nitrogen with an F and M (Model 185) CHN ca.rbon, hydro gen, and nitrogen analyzer.Duplicate or triplicate samples were run from each core sample.Total carbon was measured by dry combus tion in a Leco induction furnace and measurement of the evolved CO2 in a Leco gas analyzer (Curl, 1963), The sediment was dried, grounds, and then completely oxidized in the heat generated by the furnace. Calcium carbonate was measured by acidifying the dried ground sedi- ment with 0. 1 N HC1 and measuring the CO2 evolved,The percent by weight of organic carbon was estimated by difference between total carbon and calcium carbonate carbon. 23

Data Analysis

Data were analyzed on the Oregon State Open Shop Operating System (OS3) using the CDC 3300 Computer of the Oregon State Uni- versity Computer Center.The AIDN program developed by Dr. Scott Overton (unpublished) of the Oregon State Statistics Department was used for analysis of the grab-to-grab, station-to-station, season-to- season variation in the fauna0 Stepwise, multiple linear regression was performed itilizing the *STEP I program (Draper and Smith, 1966).The variable with the highest correlation to the data is selected at each step in the process, and those variables already entered are re-examined at every stage in the regression to determine if the contributionof a selected variable becomes insignificant and should be removed.Both the F statistic and Student's t statistic are evaluated at each step in the regression. The *FAST Factor Analysis Program of Dr. Tjeerd van Andel of the Oregon State University Department of Oceanography was used to analyze the faurial data for species groups and station groups.The *STEP Iprogram of the Oregon State University Statistics Depart- ment Program Library was used for regression analysis to study the relationship of environmental factors to particular species.The Olivetti Programma 101 desk calculator was used for converting 24 wet weights to ash free dry weights.

Some Computational Definitions and Methods

The AIDN Program calculates two measures of diversity. One of these measures is Sd2, the diversity measure proposed by Simpson

(1949),Sd2is an estimator of Simpson's 1949 index X which is de- fined for the entire population as i'E lTj2 where: = proportional value of the ith species in the population and

Z total number of species in the population. The estimator is S2 2 En. Sd =ES P.2 =E ,-11 i=1 1 i=1N2 N2

where: P. = n./N = proportion of total individuals in ith species; n. = number of individuals of ith species in the sample; N = total number of individuals in the sample; S = total number of species in the sample. The estimator ranges from 1/N to 1. The other index computed by the AIDN Program is the Shannon-Wiener information function, H.With logarithmic bases of e and 2 the 25 indices are

N H = P. logP., and . i e 1 e 1=1

N H =-P.logP.. 1 2 1 e 1=1 Lie (1968) found that the increase in H2 betweensingle samples and five pooled samples was about eight percent.The difference in H2 from five to ten samples was only four percent.He concluded that five replicate samples of 0. 1m2were sufficient to use theShannon- Wiener function H as a. valid index of diversity. The statistical comparison of samples in synecological studies has been difficult.The similarity index (SIMI) is one way ofmak- ing this comparison.Similarity can be defined as the sum of the products of the proportions of individuals in fivespecies common to two samples.

S SIM = P.P.. 1211ii2i P1. = proportion of ith species in first collection. = proportion of ith species insecond collection. S = number of species present in both collections. An index of this similarity measure can be givenby SIM I.

S1M12 SIM Limits = 0 to = (Sd1 )(Sd2) 26 This is scaled by the factor (Sd1 )(Sd2) which is the product of the square roots of Simpson? s diversity index. Grabs with no species in common have a similarity of 0, while those which have all species in common in equal proportions have a similarity of 1, The third method of comparing two samples or collections of samples is that of difference (MacArthur, 1965).The index is

[P(i 1) + P(i, 2) [P(i 1) + P(i, DIF (1,2) =-ex[ LN

2) LN P(i, 2)j [P(i 1) LN P(i, 1)] - e xp +

where: P(i, 1) = proportion of the ith species in the first collection; P(i, 2) = proportion of the ith species in the second collection;

and

N total number of species for each collection. The difference index (DIF) ranges from 1 to 2.One is total accord of the samples and 2 total difference. Levins (1968) gives two measures of niche breadth, which are measures of the dominance of a species or collection of species in a collection of samples,The measures are: 27

[P(1.J) B2 =

B5

where:

N = EP(i,j), j=' P(i, j)proportion of the ith species in the jth collection,

and N = total number of collections. The measure ranges from 1 to N. These were converted to mean niche breadth according to Mclntire and Overton (1970).

Factor Analysis

The same species subjected to the AIDN analysis of similarity and difference were also subjected to a factor analysis.Factor an- alysis gives an overall correlation among species and stations in the R and Q modes. The R Mode orders the species according to sta- tion and Q mode orders the stations according to the species compo- sition.The R Mode procedures followed were the same as those of Lie (1968).Ones were placed on the principal diagonal and the 28 species counts, X, were transformed by the equation

X* = in (X+1).

This was done to give a variance for the species that was independent of the mean (Bartlett, 1947).The positive eigenvalues and associated eigenvectors for the matrix were determined and the first six prin- cipal components rotated for interpretation.Rotation was done using the Varimax criterion of Kaiser (1958).Data for the Q Mode analy- sis were transformed by the percent range tr3nsformation according to Imbrie and van Andel (1964) by the equation

(X-Xmin) x* - . (Xmax-Xmin)x 100.

The original value is X and X* is the transformed value.This equates all data.The rare species are treated just as the most abundant species.Factor loadings of greater than ±0.were treat- ed as significant in both the R. and Q Modes of factor analysis.For a detailed explanation of the method see Lie and Kelly (1970). 29

III.RESULTS

The Benthic Environment

The 17 environmental parameters studied were sampled at each season throughout the course of the year.The results of the analysis of these 17 environmental parameters aregiven in Table 3.In this table the mean values for the four seasons are given.Complete data on the environmental parameters can be found in the Appendix.The large standard deviation in depth at station 10 is caused by relocation of the stations; it was moved from 490 meters depth to 450 meters depth to avoid a rocky area.After depth, the next six parameters are from the bottom water samples.These parameters remained fairly consistent over the range of the eight stations.Temperature and oxygen show a decrease with depth.At the time of analysis the chemistry laboratory of the Department of Oceanography at Oregon State was recording an error of up to five percent in the measurement of phosphate,For the measurements of nitrates and silicates, the error ranged up to 25 percent.This explains in part the very high standard deviation for these two nutrients.Total carbon and calcium carbonate were measured from the sediment sample taken from the Phieger corer.The values for each season are an average of three replicate subsamples.The four sediment parameters, percent total Table 3.The 17 sampled environmental parameters by station.

Environmental Parameter Station 2 6 7 8 10 15 22 23

Depth Mean 190.5 147.7 99.5 195.0 458.7 101.5 74.0 102.0 Std. Dev. 7.05 2. 63 .058 8. 12 23.79 1.91 1. 15 1. 63

Salinity %° Mean 33.94 33.59 33.36 33.92 33.95 33.72 33.33 33.68 Std. Dev. 0.07 0.42 0. 61 0. 12 1. 18 0. 26 0. 55 0. 27

Temperature °C Mean 7. 14 7. 18 8. 16 7.82 5. 57 7. 72 8.02 7.83 Std. Dev. 0.95 1.07 1.31 1.66 0.07 0.91 1.13 0.88

Oxygennil/l Mean 2.48 2.55 3.14 2.40 1.87 2.85 3.74 3.03 Std. Dev. 0.36 1.09 1.75 0. 53 0. 98 1. 19 1.65 0.85

Silicate FJ.m/1 Mean 36.25 41.7 39.50 40.75 43.25 38.50 32.75 36.25 Std. Dev. 14.41 11.59 17.60 16.64 12.66 19.14 1.88 12.12

Phosphatepm/l Mean 2.42 2.82 2.68 2.94 2.54 2.39 2.10 2.72 Std. Dev. 0.41 0.40 0.48 0.76 0.58 0. 37 0. 29 0. 34

Nitrate Jm/l Mean 24.70 25.82 24.30 25.70 28.30 23.75 19.45 24.67 Std. Dev. 8.45 5.94 10.31 10.35 6.17 9.45 11.05 5.84

°% Total Carbon Mean 1.04 1. 73 0. 63 1. 68 0.86 0.45 0. 11 1.38 Std. Dev. 0. 17 0. 19 0. 10 0. 18 0. 29 0.06 0.05 0. 10

% Calcium Mean 0.05 0.05 0.05 0.09 0.01 0.02 0.02 0.05 Carbonate Std. Dev. 0.00 0.01 0.03 0.01 0.01 0.02 0.01 0.02

% Organic Carbon Mean 0.98 1. 68 0.58 1. 59 0.85 0.43 0. 10 1. 33 Std. Dev. 0 17 0. 18 0.09 0. 18 0. 29 0.07 0.04 0. 11 Table 3.Continued.

Environmental Parameter Station 2 6 7 8 10 15 22 23

Sediment Nitrogen Mean 0. 11 0. 17 0. 62 0. 16 0.06 0. 04 0.00 0. 10 x10/gm Std. Dev. 0.01 0.02 0.01 0.01 0.02 0.00 0.00 0.01 sediment Folk and Ward Mean 4.62 6.85 4.20 7.22 2.60 2.99 1.88 6.01 Mean Particle Size Std. Dev. 0. 22 0. 34 0. 65 0.47 0.48 0. 30 0. 16 0. 57

Folk and Ward Mean 2. 77 2.86 2.31 2. 65 2. 26 1. 81 0.45 2.S8 Sorting CoefficientStd. Dev. 0.48 0. 19 0. 62 0.45 0.34 0.55 0.02 0.42

Inman Mean 3.46 5.87 3.38 6.45 2.00 2.78 1.89 4.96 Median Grain Size Std. Dev. 0.02 0.42 0.38 0.22 0.48 0. 15 0.15 0.32

% Clay by Wgt. Mean 14.74 30. 28 11. 66 35.32 7.77 7. 28 0.93 23. 60 Std. Dev. 1.49 4.56 4.63 4.62 1.93 2.15 0.86 4.62

% Silt by Wgt. Mean 16.79 66.58 22.12 62.11 13.42 8.64 0.35 48.13 Std. Dev. 1.68 4.05 1.82 2.73 9.61 3.13 0.24 2.79

% Sand by Wgt. Mean 68.46 3.13 63.72 2.57 78.81 84.07 98.72 28.26 Std. Dev. 2.70 3.01 9.39 2.19 7.91 3.53 1.06 7.39 32 carbon, percent calcium carbonate, percent organic carbon, and sediment organic nitrogen were very consistent over the four sea- sons.The standard deviation was low.The values of the standard deviations for the six textural parameters chosen are a good indica- tion of the low variability of sediment type within the station. The sample correlation matrix was calculated using the *STEP I Program. The sample correlation coefficient r was calculated ac- cording to Snedecor and Cochran (1967) by the equation

= X.X,/(X.2)(X.2) 13 1 3

where: X, = ith variable of N variables, 1 X. = jth variable of N variables, and 1 N = 17.

The null hypothesis that the population correlation coefficient P was equal to 0 was tested at the five and one percent significance levels of r.If the number of pairs of variables to be tested is N,the de- grees of freedom in this case are N- 2.The correlation coefficient matrix calculated by combining all seasons and all stations is shown in Figure 5.In this case there are 30 degrees of freedom.The value for rejecting the null hypothesis that P is equal to zero at the five percent level is 0.349. At the one percent level it is 0.449. 2 3 S 6 7 8 9 10 11 12 13 14 16 2 Salinity1 Depth 1.000 402 ...... 15 17 43 OxygenTemperature -.388-059 -.743-5821000 1000 .726 1.000 7 Nitrate65 PhosphateSilicate .225.080 156 .669.275 564 -.668-.151 670 -.741-.249- 759 1 000 .935.450 1.000 .499 1.000 98 CalciumTotal Cathon carbonate - 231 .163 .266 008 -.101 205 --.265 110 .131 084 .420 451 .180 086 1.000 615 1 000 121110 OrganicSedimentMean particle nitrogencarbon 116178 276263 - 106025 - 242255 144123 460405 199173 938998 670583 1 000 932 1 000 1413 SortingMedian grain size -.143 218 .144.173 339 .181.140 208 --.161 413 .055 213 .431.446 462 .083.102 312 .871.902 763 .749 542 .891 756 .919 755 1.000 724 1 000 .- , t . 1615 N SiltClay size -.065--.200 035 .083 214 .075 106 -.147--.133 168 -.083- 056 .060 .426 468 .099 122 .923900 .667.746 729 .918.859 888 .877.890 928 .938.980 974 .624.611 696 1.000 .945 969 1 000 .922 1.000 17 N Sand .061 .138 -.072 .175 -.090 -.471 -.120 -.927 -.717 4) -.920 -.903 -.963 4) -.678 -.965 -.956 -.986 1.000 C. C. 5. c O ., m1- 0 >. . .9.0 0 Oo cs (')C4;. o.9 ts 0. Q 0 U)Ue ,-450 - Figure 5. The correlation coefficient matrix for all stations at all (Jo -o - -ci -m _4 - seasons. 34 All variables associated with the sediment parameters, total carbon (variable no. 8) through percent sand (variable no. 17) are highly correlated at the one percent level (Figure 5).Variables eight through 16 show a high negative correlation for variable 17 (percent sand).. Thecorrelations between the last six variab1esare, of course, expected because these are reciprocal measurements of the same sample. Depth (variable no, 1) shows a high positive correlation with salinity and a high negative correlation with temperature and oxygen.Temperature, as expected, shows a high inverse correla- tion with depth and salinity, and a high positive correlation with oxygen. Of the nutrients, silicates and nitrates are positively cor- related with salinity and negatively correlated with temperature and oxygen at the 1 percent level.Phosphorus is positively correlated with silicates and nitrates and all variables in the sediment except percent sand where the correlation isnegative. In summary, salinity increases in depth while temperature and oxygen decrease. The nutrients, silicates, phosphates, and nitrates increase with increasing salinity and with decreasing temperature and oxygen.In the sediment data there is a paradox.Total carbon, calcium carbonate and organic carbon increase with percent clay and percent silt, yet they also increase with an increase in mean particle size and median grain size.This increase in grain size is usually caused by an increase in the percent sand, yet all of these parameters 35 are negatively correlatedwith percent sand.Positive increase in percent silt and clay with mean particle sizeand with median grain size could indicate an increase in large fragmentsof material mixed with fine sediments.

Station Summary

Station 2

Station 2 is at approximately 200 meters depth; thesediment is silty sand. A thin silt layer has been observedoverlying a sandy layer in the cores and in the Smith-McIntyre grabsamplesBased on a correlation coefficientmatrix forthis station(2 degrees) of freedom, the 1% level is r> 0. 990) both total carbonand organic carbon are significantly correlated at the one percentlevel with an increase in median grain size. Sediment nitrogenshows high posi- tive correlation at the five percent level withsalinity, silicates and water nitrates.

Station 6

Station 6 is at 150 meters depth; the sediment isprimarily clayey silt.Organic carbon is 1.68 percent of the sedimentby weight. 36

Station 7

Station 7 is at 100 meters depth; the sediment is silty sand. Sediment particle size is very much like that of station 2; however, the silt at station 7 does not form a layer on the top of the sand as it does at station 2, but is mixed throughout.

Station 8

Station 8 can be paired with station 6.It is at 200 meters depth and is almost identical in sediment type with station 6 (clayey silts). Like station 6 it is relatively high in percent total carbon by weight of the sediment (1.59 percent).

Station 10

Station 10 is imique among the eight stations.It is the deepest station (460 meters), and it is the only station in which glauconite composes a large percent of the sedimentIn particle size station 10 is much like stations 15 and 22.However, the sand fraction is glauconite and not beach sand. A large standard deviation in percent silt and percent sand (Table 3) is indicative of the patchiness in disC- tribution of the sediment,Of the eight stations, station 10 is lowest in temperature and oxygen concentration. 37

Station 15

Station 15, at 100 meters depth, is a silty sand station low in both organic carbon and organic nitrogen.Station 15 lies between station 6 and 8 at a shallower depth (Figure 2).This station is an example of the patchy distribution of sediments and rapid change that occurs within sediments on the Oregon shelf.

Station 22

Station 22 is at 75 meters depth. The sediment is composed of 90 percent beach sand derived from terrace deposits.The organic carbon in the sediment at 0. 1 percent byweight is the lowest of the eight stations and the organic nitrogen in the sediments cannot be detected by our methods.Station 22, with the lowest standard devi- ation (Table 3), is the most consistent in terms of sediment.The beach sand is clean and well sorted.

Station 23

Station 23 lies at a depth of 100 meters and is composed of sandy silt,This station is influenced by the Umpqua River.The standard deviations of the organic carbon and calcium carbonate parameters show a fairly high variability.Station 23 is third high- est of the eight stations in percent total carbon with 1. 38 percent 38 by weight.The composition of station 23 was the most variable of the eight stations with silt, clay, sand, and some gravel making up fractions of the sediment.

Environmental Seas onality

The water parameters derived from the Fjarlie bottle samples reflect the surface conditions on the Oregon shelf.With the increase of salinity during the summer months there is a decrease in both temperature and oxygen.This is reversed in winter with salinity being lower and temperature and oxygen being higher.The silicates, phosphates, and nitrates are highest in summer and reach low points during the winter months.The summer upwelling does not discern- ibly affect the sediment parameters studied in this investigation. However, seasonality in the sediment parameters may have been missed since replicate samples were taken only once during each season.There is no noticeable seasonality in percent total carbon, calcium carbonate, organic carbon, or organic nitrogen in the sedi- ment.Seasonal changes in sediment characteristics might be ex- pected at station 23 which is influenced by the Umpqua River, however,. none was found.See Appendix I for seasonal environmental data. 39

Sampling Results

The usual way of determining if a community is sampled ade-. quately is to plot the cumulative number of species as a function of increased number of samples.The slope at the upper end of the curve demonstrates whether the sample is adequate.If the number of species increases markedly with an increasing number of samples, more samples are needed; or if the curve is asymptotic, an adequate number of samples has been obtained.The samples are normally randomized before plotting to remove variability caused by sampling error (Longhurst, 1959; Lie, 1968).However, the samples were not taken at random in this study, but in groups of five, and following the standard procedure would not be as informative as plotting the sam- pies by seasonal groups.The cumulative number of species as a function of area sampled and the number of samples is shown in Figure 6.These samples are plotted as they were taken, the first five from the fall of 1968 and the last five from the summer of 1969.

At stations 8,6,10, and 22 the curve is fairly flat, indicating a low increase in the number of species with increased sampling; however, at stations 2,7,15, and 23 the increase is rather rapid between five and ten grabs after which it leveled off. An explanation of this increase can be better seen by looking at the percent increase in the number of species versus the increase in 40 SAMPLING AREA IN m2 0.5 1.0 1.5 2.0 0.5 1.0 1.5 2.0 I I I I I I

-40- -30- -20- -tO- STATION 2 STAflON 10 I I -40- 30 -20- -10- STATION 6 STATION 15

Ui

z STATION 7 STATION 22

I I 40-

I0 STATION 8 STATION 23

I I I I 10 I 20 IO 15 20 NUMBER OF SAMPLES

Figure 6.Cumulative number of species of molluscs, ophiuroids and cumaceans with increased sampling (20 0.1 m2 sample s). 41 number of individuals from the additional species.This information, based on the shelf mollusca, ophiuroids, and cumaceans is given in Table 4.Five of the eight stations have over 50 percent of the total number of species taken in the first five replicate samples.This agrees well with Longhurst' s data.However, two of the stations, stations 2 and 8, are considerably below the 50 percent level.Sta- tions 2,7, 8,10, and 22 also show a ten percent or greater increase in the number of species even after 15 grabs have been taken.At first glance it would appear that these communities had been inadequately sampled. However, a different picture emerges when one looks at the number of individuals from the added species.With the exception of stations 8 and 23, the number of individuals added falls rapidly after five grabs have been taken.This indicates that the species be- ing added are rare and occur in low numbers. The high percentage of individuals added from additional species at station 8 is explained by the low numerical density of fauna.Sta- tion 8 has 53 individuals of the mollusca, ophiuroids, and cumaceans per 20 grabs (i, e.2. 0m2).Species are widely scattered, and the addition of a few individuals greatly affects the percent increase. At station 23, however, the total number of species is rela- tively large.Thirty percent of the 44 percent increase in total mdi- viduals between five and ten grabs is caused by the bivalve, Macoma carlottensis.This species occurs in large numbers but with low Table 4.Percent of the total number of species and total number of individuals from the added species.

Percent of the total number of species Percent of the total number of individuals from added species Sample Station Increase 2 6 7 8 10 15 22 23 2 6 7 8 10 15 22 23

5 39.3 69.5 55.3 33.3 50.0 72. 270.0 48.8 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0

5 to 10 35.7 13.0 21.3 33.3 31.2 20.4 10.0 31.7 15.6 4.4 14.9 75.0 11.9 9.8 5.3 44.6 lOto 15 10.7 17.410.6 16.7 6.2 0.0 10.0 14.6 10.3 6.0 2.8 40.0 1.2 0.0 1.4 6.9

15 to 20 14.3 0.0 12,8 16.7 12.5 6.8 10.0 4.9 12.9 0.0 6.5 17.6 2.8 5.0 1.3 3.4

Total 28 38 47 18 16 44 20 41 no. Sp.

Total 271 265 661 53 430 538 548 614 no. mdiv. 43 frequency of occurrence; it has a clumped distribution. The surprisingly large increase in the number of species with additional grab samples shows that a complete enumeration of the infauna on the Oregon continental shelf would be difficult.However, as Longhurst (1959) has shown, this increase in number of species does not invalidate a benthic synecological study since the great ma- jority of individuals found in any group of samples comes from a few species.The small increase in the number of individuals from the additional species shows that the major species have been captured by five grab samples. Thus five 0. 1m2 samples adequatelyrepre- sent the fauna.

Species Distribution and Abundance

It is generally assumed in benthic ecology that sandy substrates have fewer species with larger numbers of individuals per species than do soft bottomed sediments (Hedgpeth, 1957).Data from the eight stations studied in this investigation do not fully support this assumption.Figure 7 which is based on the 83 species of major in- terest, plus two species of polychaetes, three species of echinoderms, and one species of nemertean illustrates this,Stations 10 and Z2, two of the predominately sand stations (see Figure 4) have the lowest number of species, yet these two stations rank fourth and fifth in number of individuals,However, station 15 which is 84 percent sand 70 60

NUMBER OF 50 40 IDENTIFIED SPECIES 30 20 I0 0 102286 223157 STATION

700

600

NUMBER OF 500 INDIVIDUALS 400 300-

200

100

8 6 2 102215237 STATION

Figure 7.Total number of identified species and individuals of molluscs, ophiuroids and cumaceans per two meters2. 45 is second in total number of identified species and third in total num- ber of individuals. A comparison between rankings based on species and on individuals demonstrates that only the sand stations, 10 and 22 shift rank in relation to the others. The distribution of species among the major taxonomic groups at each station is shown in Figure 8.Over 90 percent of the number of individuals at each station are either polychaetes or molluscs. Crustaceans and echinoderms comprise a small part by number of the total fauna. Animals not belonging to these four major groups make up less than one percent of the total fauna by numbers. Except at station 8, the molluscs can be seen to make up a significant per- cent of the total number of individuals at each station.The number of individuals in each of the major groups by station is shown in Table 5. A total of 82 species of molluscs, cumaceans, and ophiuroids were identified in this study.There were four species of scapho- pods, 31 species of bivalves, 30 species of gastropods, ten species of cumaceans, and eight species of ophiuroids.The identified species are listed by family and station in the Appendix. For statistical an- alysis each of these plus six additional species and one group for decalcified individuals, of which there were few, were given a code number (Table 6). Of the three groups under major consideration the molluscs 46

MISCELLANEOUS 1000 POLYCHAETA

ECHINODERMATA 900 CRUSTACEA MOLLUSCA 800

N E 600-

500-

400-

300- z

810622715 STATION

Figure 8.Distribution of the total number of individuals by major taxa. Table 5.Number of individuals permeter2by major taxa.

Station Pelecypoda Scaphopoda Aplacophora Crustacea Echinodermata Polychaeta Miscellaneous Total

2 99 13 23 3 36 13 266 6 459

6 78 31 22 10 12 7 266 4 430

7 137 69 114 8 32 17 558 13 948

8 12 3 5 9 17 9 188 4 247

10 203 4 4 15 8 106 3 343

15 165 70 23 30 14 672 14 988

22 179 79 15 33 3 230 4 543

23 152 44 96 2 7 19 495 6 821 48

Table 6.Species code list for all molluscs, ophhzroids and cumace1ps.

Code Number

1 Aila castrensis Hinds 2 Nucula tenuis Montagu 3 Nuculana aust Oldroyd 4 Nuculana minuta Fabricius 5 Yoldia ensifera Dali 6 Yoldia thraciaeformis Storer 7 Yoldia cecinella Dali 8 Huxleyi a minuta Dali 9 Crenella discussata Montagu 10 Megacrenella coiumbiana Dali 11 Musculus nigra Gray 12 Cardita ventricosa Gould 13 Pseudopythina "A" 14 Lucinoma annulata Reeve 15 Adontorhina cyclia Berry 16 Axinopsida serricata Carpenter 17 Thyasira bisecta Conrad 18 Thyasira gouldii Philippi 19 Neniocardium richardsoni Whjteaves 20 Compsomyax subdiaphana Carpenter 21 Psephidia lordi Baird 22 Macoma carlottensis Whiteaves 23 Macoina elimata Duhfll & Coan 24 Tellina carpenteri Dali 25 Tellina salmonea Carpenter 26 Pandora filosa Carpenter 27 Lyonsia pugettensis Dali 28 Cardioniya pectinata Carpenter 29 Cardiomya planetica Dali 30 Odontogena borealis Cowen 31 Thracia trapezoides Conrad 32 Decalcified bivalve 33 Solariella varicosa Mighels and Adams 34 Solatiella nuda Dali 35 Balcis sp. 36 Epitoniurn caamanoi Dali and Bartsch 37 Epitonium tinctum Carpenter 38 Epitonium Dali 39 Tachyrhyncus lacteolus Carpenter 40 Bittiuin mintm Carpenter 41 Polinices pallidus Broderip and Sowerby 42 Boreotrophon dalli Kobeit 43 Fxilioidea rectirostris Carpenter 44 Mohnia exquisita Dali 45 Neptunea liratus Ginelin 46 Mitrella gouldi Carpenter 49

Table 6.Continued.

Code Number

47 Nas!arius fossatus Gould 48 Nassarius mendicus Gould 49 baetica Carpenter 50 Olivella biplicata Sowerby 51 Oenopota sp. 52 Ophiodermella rhines Dall 53 Ophiodermelia incisa Carpenter 54 Rectiplanes thalaea Dali 55 Mangelia sp. 56 Odostpmia sp. 57 Turbonilla pedroana Dali and Bartsch 58 Turbonilla aur antia Carpenter 59 Acteocina culcitella Gould 60 Acteocina eximia Baird 61 Cylichna attonsa Carpenter 62 Acteon punctocaeiatus 63 Unioplus macraspis 64 Aniphiodia urtica 65 Op}iiura sarsi 66 Ophiura iutkeni 67 Axiognathts pugetana 68 Unioplus euryaspis 69 An-iphiurjdae 70 Ophiura sp. 71 Campylaspis rubicunda 72 Ewlorella pacifica 73 Leucon longirostris 74 Hem ilamprocaliforniensis 75 Diastylis paraspinulosa 76 Eudoreila pellucida 77 Diastylisdalli 78 Diastylis bidentata 79 Colurostylis occidenta].is 80 Diastylis "A" 81 Dentnliuin rectius 82 Dentalium pretiosum 83 Cadulus stearnsii 84 Cadulus californicus 85 Sternaspis scutata Renier 86 Travisia brevis Moore 87 Pentamera populifera Stimpson 88 Brisaster bretifronsAgassiz 89 Allocentrotus fragilis Jackson 90 Nemertine sp A 50 play a predominant part0The numbers of individuals of cumaceans and ophiuroids are quite low; therefore, the molluscs were chosen to illustrate the faunal differences between stations.The number of individu3ls permeter2of the most important species and their per cent occurrence in each station are demonstrated in Figures 9and

10.The importance of a particular species at a given station was determined by ranking the species within that station by number of individuals permeter2and by percent occurrence within the 20 grabs taken.The ranking of each species in percent occurrence was multi plied by the correspondingranking to give an importance value. Only those species that have a sum total of more than six individuals for the 20 grabs were considered. The general similarity of stations 2 and 6, and stations 23,7, and 15 can be seen in the graphs.Station 10 and station 22 are unlike any of the other stations.The low number of species at station 8 is caused by the rare species that were too low in numbers to be used. The graphs demonstrate that a very small amount of clumping occurs,Those species that are high in numbers generally have a high percent occurrence.Fifty percent occurrence means that a species appeared in ten of the 20 grabs taken at that station.With this method a clamped distribution is quite evident, At station 10, species no, 9, Crenella discussata, has over 200 individuals and yet occurs in only 30 percent of the grabs. At station 23, species 51 STATION 8 320 ISO- 80- 40- NUMBER OF INDIVIDUALS PER2m2 81 6 7 9 842446 8 15 830223356 25 %50 FREQUENCY OF 75- OCCURRENCE 100

STATION 10 320 160- 80 40- NUMBER OF 20 (0 flu INDIVIDUALS 5 PER2m2 81 16 7 9 8424 46 18 (5 8 30223356 25 %50- FREQUENCY OF 75- OCCURRENCE 100 STATION 2 320- (60-on- 40 NUMBER OF 20 I0 INDIVIDUALS 5 PER2m2 930842224334656

2 %5o FREQUENCY OF 75 OCCURRENCE 100

STATION 6 320 60- 80 40 NUMBER OF 20 10 INDIVIDUALS 5 PER2m2 811816 30842224334656

25 %50 FREQUENCY OF 75 F OCCURRENCE 100

Figure 9.Distribution of major molluscan species at stations 8, 10, Z, and 6. 52

NUMBER OF INDIVIDUALS PER 2rn2

FREQUENCY OF OCCURRENCE

STATION 23 320 16 80 40 NUMBER OF 20 I0 INDIVIDUALS 5 PER 2m2 JhtTIP!!!!181 16 7 9 842446 13 25 12 61 49 40 23 60 48 4 18 IS $ 30 2233 56 21 I 825 2083 14 10 II 41 25 %50 FREQUENCY OF 75 OCCURRENCE 100 Ullilli101 STATION 7 320 16 80 40 NUMBER OF 20 10 INDIVIDUALS 5 "I PER 2m2 81 16 7 9 8424 46 13 2.5 49 40 2.3 12 61 60 484 18 IS 8 30223356 21 I 82 5 20 83 14 JO II 41 2 % 50 FREQUENCY OF 75 OCCURRENCE 100

STATION 15 320 160 80 NUMBER OF 40-20 I0 INDIVIDUALS .5 PER 2m2 81 16 7 9 8424 46 13 25 12 61 60 1$ 49 40 23 48 4 15 8 30223356 21 I 825 20 83 14 JO II 41 25 % 50 'p 'U FREQUENCY OF 75... OCCURRENCE 100 I

Figure 10. Distribution of major molluscan species at stations 22, 23,7, and 15. 53 no. 22, Macoma carlottensis, has over 100 individuals that occur in only 25 percent of the grabs. Mitrella gouldi, species no. 46, is an abundant, wide-spread species at stations 15, 8,10,2, and 6.These five stations cover a wide range of sediment types from nearly 100 percent silt at station 6 to nearly 100 percent sand at station 22,The factors determining the distribution of this gastropod seem to be something other than sediment type.

In - Station Variation

The statistical parameters provided by the AIDN Program allow us to measure the within station variation of the grabs from season to season.Because of the limitations of the program, the five grabs from each season were lumped together and analyzed as a block. This is an acceptable procedure because five grabs were originally taken as a block on each cruise to provide a reasonable estimate of the fauna at a particular station for each season. One measure of within-station variation is the similarity index, SIM I.The mean and range of the six pairs of samples that can be compared among the four seasons at each station are illustrated in Figure 11,This is not a comparison of station-to-station variation, but only within station variation.At six of the stations the similarity is high and the range is relatively small; this indi,cates that there is 54

1.0- 0.9- 0.8- 0.7- 0.6- 0.5- SIMI 0.4 03- 02-

0.1- 00-

I I I I I 267810152223I

STATION

Figure 11. Within station variation of similarity index (SIMI) during the sampling period. 55 little seasonal variation in either the species presentor the propor- tion of individuals of each species within these stations,However, stations 8 and 10 show a very wide range of variation and an inter- mediate similarity. As demonstrated previously, this variation at station 8 can be explained on the basis of a few specimens of arela- tively large number of species.The large variation at station 10 is an indication of the clumping of the very abundant bivalve, Crenella discussata, Two blocks of samples which both have Crenella discus- sata present show a high similarity while those that do not have C. discus sata show a low similarity with those that do. These variations would be nullified if only presence-absence data were used rather than proportions of individuals, however. Presence-absence is thus less realistic and represents a loss of data, We are not only interested inwhat speciesoccur but what their abundance is when they do occur. A different manner of looking at the same fauna is given by difference index (DIF).This index varies between one and two.The greater the difference between the samples, the higher the number, The in-station variation in the difference index is plotted in Figure 12. As can be seen again the same picture that emerged with the similar- ity index is seen here.Station 8 and station 10 show a high difference and much variation within station.All six other stations show a very. low difference and very low station variation. 56

2.0

DIFFERENCE INDEX

I I -I I I I 2 67810152223

STATION

Figure 1 2. Witliing stationvariation of difference index (DIF) during the sampling period. 57

The in-station variation in diversitySd2can be seen in Figure

13,Station 8 which has very low similarity and very high difference within station shows a high diversity, i.e. a low number of individ- uals per high number of species.The species at sta,tion 8 are also not evenly distributed by season. The two sand stations, station 10 and station 22, show the lowest diversity and the greatest range within station.Because of the low number of species at these two stations,Sd2changes considerably if one species changes its pro- portions in a given block of samples. The clumping at station 10 would cause part of the variation here, while the placement of the mean at station 22 indicates that it is only one set of samplesthat has caused a large amount of variation.This was due to the low proportion of species no. 13, Pseudophythina hA,during the winter sampling period.Stations 10 and 22 have a generally low diversity (Figure 13),Station-to-station diversity will be discussed in alater section.The variation within station for difference similarity, and Simpson! s Sd2 within station are listed inTable 7. A fourth and final method of looking at in-station variation is that of the weighted mean niche breadth. Summed for all the species in the station it is a measure of endemism. Low niche breadth for a group of samples means that the species found in those samples are found primarly in those samples, and few or none are found else- where.Table 8 gives the means of Levin s two measures of niche 58

'.0- 0.9- 0.8- 0.7- 0.6- 0.5- Sd2 0.4- 03- 0.2- 0.1- 0.0-

I I I I I I I I 267 810152223

STATION

Figure 1 3. Within station variation of Sd2 duringthe sampling period. 59 2 Table 7.Within station comparisons of DIF, SIMI, and Sd

2 6 7 8 Dif 10 15 22 23

1.11 1.10 1.11 1.63 .1.47 1.11 1.08 1.26 1.14 1.08 1.10 1.50 1.54 1.12 1.05 1.11 1.20 1.21 1.10 1.45 1.05 1.13 1. 12 1.21 1.15 1.11 1.08 1.53 1.05 1.13 1.04 1.24 1.17 1.21 1.10 1.36 1.41 1.13 1. 10 1. 20 1.21 1.20 1.14 1.25 1.47 1.16 1.07 1.12

6.98 6.91 6.63 8.72 Total 7.99 6.78 6.46 7.14

1.16 1.15 1.10 1.45 i. 1.33 1.13 1.08 1.19 1.11- 1.08- 1.08- 1.25- 1.05- 1.11- 1.04- 1.11- 1.21 1.21 1.14 1.63Range 1.54 1.16 1.12 1.2

2 6 7 8 SIM I 10 15 22 23

.872 .841 .910 .306 .198 .922 .911 .658 .916 .938 .904 .476 .158 .878 .951 .952 .695 .808 .936 .537 .992 .897 .834 .876 .849 .838 .979 .375 .972 .922 .928 .673 .840 .788 .897 .538 .290 .850 939 .686 .743 .805 .882 .677 .258 .809 .829 .936

4.915 5.018 5.508 2.909 Total 2.868 5.278 5.392 4.781

.819 .836 .918 .485 R .478 .880 .899 .797 695- 788- 882- 306- . 158- .809- 829- 658- 916 938 936 677 Range.992 .922 951 952

2 Sd .1463 .1777 .1287 .1875 .5845 .1058 .2052 .1725 .1368 .2699 .1299 .1250 .3271 .0871 .2432 .1719 .1748 .1829 .1894 .1800 .4832 .1094 .2285 .1658 .1181 .2851 .1068 .1211 .4626 .0999 .4501 .1333

5760 9156 .5548 . 6136 Total L 8574 .4022 1. 1270 . 6435

1440 . 2289 . 1387 .1534 .4643 .1005 .2817 .1609

118- . 178- . 107- . 121- .327- . 087- . 205- . 133- 175 . 285 . 189 187Range 584 109 450 172 60

Table 8. Two measures of weighted mean niche breadth (Levins, 1968) within station by season. Station Season R(2) R (5)

2 Fall 13.0196 11.0321 Winter 12. 3893 10.5026 Spring 14.2168 12. 1772 Summer 10.1254 8. 4326

6 Fall 14.4556 la.3017 Winter 14. 4762 12. 4625 Spring 15.9335 13.8026 Summer 14. 7282 12.4691

Fall 14.3696 12.7228 Winter 13. 4880 11. 9357 Spring 15.1.327 13.4969 Summer 13.4443 11.9376

8 Fall 12.0808 10.5194 Winter 11.8173 10.0470 Spring 10.9520 9.4437 Summer 12. 2766 10. 3883

10 Fall 2. 6490 2.4139 Winter 5.3696 4,1957 Spring 5.6172 4. 2377 Summer 3.3581 2.9081

15 Fall 12. 3411 10. 6762 Winter 12. 0349 10.3575 Spring 12. 6427 10.9087 Summer 12.4046 10.7166

22 Fall 5.5663 4.7890 Winter 5. 7434 4. 8590 Spring 6. 6199 5. 6814 Summer 6.5033 5.4509

23 Fall 13. 4888 11. 8995 Winter 12.0883 10. 1627 Spring 13.9121 12.3282 Summer 13.9517 12. 1116 61 breadth (McJntire and Overton, 1970).The species that occur at station 10 and station 22, the two sand stations, are found here and are not found at other stations in any great proportion.Conversely, the other six stations, the sandy silt and silty sands, show a wide interchange of species.Within, each station there is little seasonal variation.The greatest variation comes at station: 10; this is again due to clumping.

Station- to- Station Variation

Station-to-station variation has been analyzed by the same methods used for variation within the station.To study station-to- station relationships the seasons within a station are summed and the various indices computed on these sums. Similarity indices on a station-to-station basis are listed in Table 9.Station 2 is most simi- lar to stations 6 and 8; station 6, likewise, is most similar to 8 and to 2.Station 8 is closely related to 7 which is in turn closely related to stations 15 and 23.Stations 10 and 22 are not closely related to any other. The difference index points out a similar pattern (Table 10). Stations 2,6, and 8 are closely allied and station 7 is much like 6 and 8.Station 15 and 23 arequite close and these are, in turn, closest to station 7.Again, 10 and 22 are very different from all the others as indicated by the high values; they are also very different Table 9.Similarity index (SIMI) between stations.

Station 2 6 7 8 10 15 22 23

2 1.000

6 .788 1.000

7 .569 .517 1.000

8 .777 .726 .771 1.000

10 .002 .020 .008 .066 1.000

15 .314 .451 .557 .574 .139 1.000

22 .011 .079 .072 .043 .000 .173 1.000

23 .488 .387 .885 .606 .007 .384 .064 1.000 Table 10.Difference index (DIF) between stations.

Station 10 15 22 23

2 1.000

6 1.216 1.000

7 1.320 1.313 1.000

8 1.222 1.225 1.324 1.000

10 1.943 1.922 1.910 1.859 1.000

15 1.427 1.383 1.1w 1.374 1.766 1.000

22 1.918 1.783 1.778 1.867 1.991 1.612 1.000

23 1.374 1.413 1.109 1.443 1.922 1. 284 1.805 1.000 64 from each other.The differences between stations are much larger than they are within a station, Another widely used measure of diversity is the Shannon Weiner function H.Simpson's diversity indexSd2and the Shannon-Weiner function are illustrated in Figure 14.The sand stations, 10 and e 22, are the lowest in diversity.It is surprising, however, that the station with the greatest diversity is station 15, because this station has 84 percent sand.The Shannon-Weiner function shows separation of the station groups as shown by similarity and difference.Station

22 has low diversity; 7,15, and 23 have high diversity, while 2,6, and 8 are intermediate.

Simpson'sSd2gives a very similar picture with the exception that station 8 becomes second highest in total diversity.Both indices utilize the proportions of individuals and treat the data in a similar fashion. The last measure of station-to-station difference, the weighted mean niche breadth (Table 11), delineates stations 10 and 22 as nar- row and different from any of the other stations.This result is con- sistent with the other measures for station-to-station differences. The station-to-station relationships for similarity and for differ- ence are summarized in Figure 15,The most similarity and the least difference between two stations are indicated.Stations 2, 6, and 8 are a station group; 7 is closest to 2; 15 and 23 are close, and 10 and 102262 8237 15 STATION Figure 1 4.Comparison of the Shannon-Weiner function, He and Simpson's diversity index, Sdfor all stations. 66

MOST SIMILARITY

I0

LEAST DIFFERENCE

I0

Figure 15. A two dimensional representationof the relationships between stations based on the greatest value of SIMI between stations (upper) and the lowest value of DIF (bottom). 67

Table 11. Two measures of weighted mean niche breadth (Levins, 1968) summed for aU seasons within a. station.

Station B2 B5

2 3. 8895 3. 3520

6 4. 4372 3.8338

7 4. 2196 3.7775

8 3.8758 3. 3702

10 1.5928 1. 3676

15 3. 7835 3.2931

22 1. 8015 1.5734

23 4. 1035 3.6329 68 22 are far removed from any of the others.Using the difference in- dex, 2, 6, and 8 are identical in difference.Stations 2,6, and 8 are also equidistant from station 7 to form an isosceles triangle.Sta- tions 7, 23, and 15 are not very different from each other while sta- tions 10 and 22 are very different from any of the others and from each other.

Season-to-Season Variation

By summing separately the difference and similarity indices for eight stations, we can compare seasons at all eight stations to each other (Tablet 1 1).There is no significant difference between seasons in the samples analyzed.Likewise, the similarity index shows a very high similarity between all seasors. Simpson's diver- sity index, the Shannon-Weiner function H., and two measures of niche breadth for each of the seasons demonstrate that there is no significant seasonal variation in the fauna as awhole (Table 13). To investigate the possibility that the large number of species from the three groups was masking any seasonal variation in the numbers of individuals per species, the entire set of data was run over again for the molluscs only,The results using just the mol- luscs are almost identical to those using the molluscs, cumaceans, and ophiuroids,The full set of data on molluscs can be found in the Appendix.In addition, the results are also almost identical 69

Table 12.Difference index (DIF) and similarity index (SIMI) between seasons summed for all stations.

Fall Winter Summer Spring

Fall 1.000 Winter 1.085 1.000 Summer 1.088 1.057 1.000 Spring 1.064 1.110 1.101 1.000

Similarity Index Fall Winter Summer Spring

Fall 1.000 Winter .875 1.000 Summer .877 .935 1.000 Spring .920 .829 .829 1.000

Table 13.Seasonal variations in Sd2, He, B2, B5 summed for all station.

Season B2 I3

Fall ..0660 3. 195 3.522 3.392 Winter .0587 3.257 3.634 3.451 Summer .0740 3.073' 3.706 3.549 Spring .0602 3. 235 3.512 3.371 70 analyzing three grabs instead of five indicating that 0. 3m2is a sufficient area to sample seasonally. Seasonal variation can also be studied by using numbers of individuals (Figure 16).The very large number of bivalves in the fall at station 10 and winter at station 23 is due to the clumped spe- cies Crenella discussata and Macoma cariottensis.At station 2 the fall peak is again caused by Macoma carlottensis.Although there is some variation in number, no pattern can be seen.

Standing Stock

This investigation, like many others in benthic synecology, has as one of its primary aims the evaluation of the standing stock as a food source for demersal fishes.Biomass was divided into edible and unedible fractions to better assess food supply available to the fishes.Data from preliminary stomach analysis of the Dover Sole, Microstomus pacificas at Sea Grant stations 2 and 6 aided in deter- mination of food fractions.The edible portion of the infauna consists of all the infauna minus the echiurids, holothurians, echinoids, and burrowing anemonesThe total wet weight of the edible portion, by season, is given in Figure 17. Station 10 ranks fifth in number of individuals; but it supports the lowest total wet weight bibmass of fauna. At this sand station, most of the animals are small, and the large number of bivalves 71 100 100- eo- 60- 40- 40- 20- 20-

STATION 2 60 60 STATION 7 40- 40- 20 20- E .0 120 -ISTATION 6 STATION 8 0 120 I00 100- 80 80- 60- 40 40 oU- 20 20- STATION 15 STATION 22 I8Ô- 180 160 160- Z 140- 140-

120- KEY 120- F= FALL 100- W WINTER 100- 6p SPRING 80- S SUMMER 80- 60- 60- 40-- 20 20 STATION 10 STATION 23 WSpS F WSpS

Figure 16. Seasonal variation in the number of bivalves for all stations. 72 105- 100- 95- SUMMER 90 SPRING 85 WINTER FALL 80 75- 70 65- WET 60- WEIGHT 55- qm/m2 50- 45- 40- 35 30 25 20

15 I0 5

STATION2 6 7 8 1015 2

Figure 17. Wet weight ingrams permeter2of the edible fraction of the biomass within all stations. 73 present are very thin-shelled.Because of the very large fraction of molluscs in the sample,. station 22 ranks fourthin number of individ- uals but supports the greatest biomass.This is due to Acila castren- sis, the dominant bivalve at this station, which is quite large and heavy-bodied. No seasonal pattern could be demonstrated in the wet weight biomass. The ash-free dry weight standing stock (Figure 1.2) is much the same as that of wet weight.However station 10 ranks relatively higher than stations 2, 6, and 8 due to the large molluscan fraction. Molluscs have a higher conversion to ash free dry weight than echin- oderms.Again, there seems to be no seasonal pattern. Total ash free dry weight of all animals contained in the grabs is shown by major taxa in Figure 19 and Table 1.4.There is a very large increase in ash free dry weight at stations 2. and 8 due to the presence of a large echinoid, Brisaster latifrons. A large propor- tion of the ash free dry weight is comprised of molluscs at stations

7,15, 22, and 23.The closely related stations 2,., 6, and 8 have a large fraction of their biomass as ecbinoderms.The echinoderm fraction appears large at station 10 becuase the large number of bi- valves are very small.The presence of a few large echinoids great ly influences the biomass. The edible fraction of the biomass (wet weight) illustrates the lack of seasonal variation in the biomass (Figure 20),There are SUM

50 pS4 4.5 oR(\V

20 t.5

t.O .5 - _ t0152' shlrp'1Ot4

'jgUre iS. £reedr' ight gratper 2oi the £raCtI'fl the a11tat10t 75 4.5-

MISCELLANEOUS 4.0- POLYCHAETA

ECHINODERMATA CRUSTACEA 3.5- MOLLUSCA N E 3.0 =_ w 2.5- >- I

2.0-

lJ- = Ci) 1.5-

E

1.0-

0.5- ii;i

267 8 10 15 STATION

Figure 19. Total ash free dryweight in grams per meter2 of the major taxa by station. Table 14. Ash free dry weight in grams permeter2 of themaj or taxa by station.

Station Polycha eta Echinodermata Crus.cea Mollusca Miscellaneous Total

2 0.1395 0. 7858 0. 0043 0. 1567 0.0368 1. 1231

6 0.1712 2.2863 0.0083 0. 2018 0. 2214 2. 8889

7 0.1745 1. 0214 0. 0046 1.0990 0. 2043 2. 5038

8 0.1879 2. 4694 0.0030 0. 0667 0. 0710 2. 7977

10 0.0345 0. 7807 0.0046 0.0476 0, 038 1 0. 9055

15 0.0814 0. 9337 0.0041 2. 0870 0. 0572 3. 2634

22 0.0449 0.0070 0. 0059 4.4963 0. 0174 4. 5716

23 0. 3023 0. 3061 0.0011 1. 7433 0. 2529 2. 6058 77

40 STATION 2 40-STATION 8 30 30- 20 20-

10 10-

40 STATION 6 40- STATION 10 30 30- 20 20- IO 10- 10

STATION 7 40- STATION IS 30 30- 20 20 10 I- LU 70 STATION 22 70-STATION 23 60 60- 50 50- 40 40- 30 30- P 20 20' I0 10 FWSp$ KEY F WSp S F :FALL WWINTER Sp:SPRING S SUMMER

Figure 20. Seasonal variation ingrams permeter2 of thewet weight of the edible fraction of the biomass. 78 anomalies in the seasonal pattern that can be explained by the pres- ence of some large animal in one ortwo grabs. At station 10 the large winter weight was caused by three large ophiuroids, Ophiura sarsi, At station 15 the winter variation is caused by two large bivalves of the species Cardita ventricosa.The summer peak at station 23 is due to three large polychaetes of the species Travisia brevis.These anomalies are also demonstrated by ash free dry weight (Figure 21). Nevertheless, there seems to be no consistent seasonal pattern in either the wet weight or the ash free dry weight for biomass.

Factor Analysis

R Mode

In the It mode of factor analysis 15 factors were originally ex- tracted which explained 75. 8 percent of the total variance. Of these 15 factors the first six accounted for 46. 2 percent of the total vari- ance.Beyond the sixth factor there was a gradual decrease in the percent variance explained by each of the succeeding factors, all contributed less than four percent of the total variance.Table 15 shows the eigenvalue, the percent of the total variance explained, and the cumulative percent of the total variance for each of the 15 factors. The first six factors were chosen for rotation and interpretation. Four distinct species groups resulted from varimax rotation (Table 16). 79 2.0 STATION 2 2.0 -STATION 8

tO- 1.0-

2.0 STATION 6 2.0 -STATION 10

1.0- -'In -1- STATION 7 20 STATION 15

1.0

STATION 22 30 -STATION 23

2.0-

1.0 1.0

KEY F WSp S F: FALL F W Sp S W WINTER Sp: SPRING S SUMMER

Figure 21. Seasonal variation ingrams permeter2 of theash free dry weight of the edible fraction of the biomass. Table 15. The 15 factors extracted by R mode factor analysis: eigenvalue, percent total variation explained, and cumulation percent total variation by factor.

Factor Figenvalue Percent Cumulative Percent

1 4.69543 13.41551 13.41551

2 3. 55251 10. 15002 23. 56553

3 2.50754 7. 16440 30. 72992

4 2. 23332 6.38093 37. 11085

5 1. 85996 5.31417 42. 42502 6 1. 67686 4. 79104 47. 21606

7 1.46045 4.17271 51.38877

8 1.37969 3.94198 55.33075

9 1. 17003 3.34294 58. 67369 10 1. 11976 3. 19932 61.87301

11 1.07019 3. 05769 64, 93070

12 1.01390 2.89684 67.82755

13 .98118 2.80337 70.63092

14 .95863 2. 73894 73.36985

15 .85493 2.44266 75.81251 Table 16. The four species groups extracted by R mode factor analysis.

Code Species FACTORS I II III IV V VI

10 Megacrenella columbiana 0,57 0. 19 0. 05 -0. 14 -0.47 0. 13 18 Thyasira gouldii 0. 66 0. 16 -.04 -0.06 -0,34 -0.09 40 Bittium minutuni 0.52 0. 20 -0.03 -0. 14 0. 25 -0.30 83 Cadulus ste arnsii 0.52 0.30 -0. 29 -0.09 0.16 0,07

01 Acila castrensis -0.47 0. 68 0.18 -0.24 -0.08 -0.06 25 Tellina salmonea 0.44 0, 60 0.26 -0.21 -0.07 -0,09 46 Mitrella gouldi 0.39 0. 55 -0.08 0.02 -0,26 0,09 49 Olivella baetica -0.43 0. 69 0.12 -0. 26 -0. 13 -0. 05 82 Dentaliwn pretiosum -0.38 0.21 -0.21 -0.08 0.10

12 Cardita ventricosa 0. 29 0, 22 -0.50 0.05 0. 24 0. 32 11 Musclus niger -0.03 -0.05 -0.57 0, 26 -0.27 -0. 21 22 Macoma carlottensis 0.02 0. 23 -0.69 -0.09 0.22 0.08 81 Dentalium rectius 0. 34 0. 17 -0.58 0,18 0.37 -0.14

08 Huxleyia minuta 0.13 -0.39 0.13 -0. 65 -0,10 0.17 09 Crenella discussata -0.02 -0.32 -0.05 -0. 65 0.01 -0.06 30 Odontogena borealis -0.04 -0.47 0.08 -0.72 0.06 0.07 84 Cadulus californicu$ -0.05 -0. 28 -0. 18 -0.18 -0.22 82 Species Group I,Species group i is most clearly represented at station 7.It is only at this station that all four of these species oc cur in any abundance, These species occur lessabundantly at station 15 and station 23,These three stations are the only stations at which these four species occur. Species Group II.Species group II is clearly representative of station 22.Three of the five species, Dentalium pretiosum, Olivella baetica, and Tellina salmonea are found only in station 22.Acila castrensis is found at stations 7,15, and 23 also, but in much lower numbers. Mitrella gouldi is very widely scattered and abundant at four stations (see Figures 9 and 10),It has, however, a very high frequency of occurrence at station 22 (85%) and is quite abundant with 23 individuals permeter2.These five species are the five most abundant species at station 22, pecies Group III.Species group III is most nearly representa- tive of station 23, but like those species that loaded on factor1, these species also are found to a lesser degree at stations 7 and 15.The three species, Cardita ventricosa, Macoma carlottensis, and Deri talium rectius, are found in abundance at station 23, while thefourth species, Musculus niger, has only six individuals at station 23.At the very similar station 15 there ar e ten individuals present; however, station 15 has very low numbers of species 22, Macoma carlottensis. Stations 15 and 23 are very close to station 7 although they differ 83 slightly from station 7 in some of their components.

Species Group IV.Threeof the fourspeciesin this group, Crenella dis cus sata, Odontogena borealis, and Cadulus californicu.s, are found only at station 10.The fourth species, Haxleyia minuta, although found at other stations, is abundant only at station 10.The separateness of this group is seen by the low loadings, or their re- versed sign within a loading, to all other factorsSpecies group II has high positive loadings in factor 2, while species group IV has high negative loadings in factor 2, Stations 2, 6, and 8 had no species group particular to them possibly because of the low numbers of species and the relatively low number of individuals.Their dominant fauna, Axinopsida sericata, Thyasira gouldi, and Dentalium rectius, also occur to a lesser de- gree at other stations.If a wider range of taxonomic material was used in the R mode factor analysis, perbaps a species group ascrib- able to these stations would become evident,

Q-Mode

The Q mode of factor analysis was origimafly run for 160 sam- ples utilizing 89 species (Appendix 2) in combination with the 17 en- vironmental variables,The results, using all the species and the environmental factors, explained 86. 8 percent of the total variance; however, 55. 7 percent was explained by the first factor and an 84 additional nine percent by the second factor.Because this was im- possible to interpret meaningfully, the Q mode was run again minus the environmental factors to make it conform to the study of Lie and Kelly (1970). Four factors were selected for rotation and interpre- tation.These four factors explained 62.3 percent of the total vari- ance; 23. 3 percent of this total variance was still ascribable to factor

1.However, the results were amenable to interpretation. A significant number of grabs, greater than 50 percent at each station, show a high loading to the first factor at stations 2,7,15, and 23 (Table 17). A number of grabs within these stations also have significant loading to factor 2.These two factors are thought to repre- sent the silty-sand, sandy-silt stations that lie between stations 22 and 10.The only station that has shown a significant loading at all to factor number 3 is station 22,At this station 75 percent of the grabs show a high loading to factor number 3. No other sample at any sta- tion shows a high loading to this factor. Factor number 4 has significant loadings only within station 10, Sixty percent of the grabs taken at station 10 have a factor loading greater than ±0. 5 to this factor. Only one other grab at the other seven stations shows a significant loading to this factor; this is grab number 65 at station 8 which shows aloading of 0. 58 to factor number

4.There is no explanation for this except only three individuals of the 89 species studied were found in this sample. Only 40 percent of the 85

Table 17.Results of9mode factor analysis with four factors rotated.

Station Sample Number I II III IV

2 3 .53 .78 -0.11 -0.09 5 , 54 , 55 -0. 13 -0. 10 6 .53 -0.31 -0.33 -0.31 7 .52 .44 -0. 14 -0.09 9 53 .36 -0.01 -0. 15 10 .57 .05 -0.31 -0.20 11 .51 -0.29 -0.24 -0.22 12 .59 .25 -0.32 -0.30 13 . 52 -0.35 -0. 25 -0. 27 16 .54 .77 -0.12 -0.07 18 .53 .24 -0.18 -0.08 Station 6 23 .52 -0.38 -0. 27 -0.32 26 .53 -0.48 -0.29 -0.27 28 . 56 -0. 37 -0. 26 -0. 16 30 .51 0.23 -0.16 -0.18 31 .56 -0.39 -0.31 -0.36 32 .53 -0. 30 -0.09 -0. 23 34 .53 -0.30 -0. 24 -0. 28 36 .56 -0. 61 -0. 17 -0.09 Station 7 41 .55 .42 -0.16 -0.14 42 .58 .54 -0.19 -0.15 43 .58 .47 -0.20 -0.07 44 .54 -0.09 -0.14 -0.03 45 .52 .00 -0.11 -0.06 46 .58 .40 -0. 18 -0. 18 47 .55 .39 0. 14 -0.12 50 .55 .40 -0. 17 -0.05 51 .58 .30 -0.20 -0.09 52 . 62 .41 0. 23 -0.06 53 54 -0.07 -0. 17 -0. 10 54 56 -0.08 -0. 19 -0. 20 55 .58 .48 -0.20 -0.16 57 . 51 -0. 16 -0. 15 -0.09 58 .57 40 -0. 15 -0.04 59 .55 .32 -0.10 -0.09 Station 8 65 .42 -0.08 .07 .58 69 .54 .78 -0.12 -0.09 78 .50 .36 -0. 10 -0.03 86

Table 17.Continued.

Station Sample Number I II III IV

10 83 .45 -0.16 .00 .53 86 .47 -0. 14 .03 .73 87 .43 -0.08 .09 .73 88 .42 -0.07 .08 .58 89 .46 -0.01 .03 .63 90 .43 -0.07 .09 .70 91 .42 -0.07 .08 .62 92 .45 .01 .02 .53 93 .41 -0.08 .09 .62 94 44 -0.08 .09 .84 95 .44 -0.07 .08 .63 98 .40 -0.08 .08 .50 Station 15 101 .59 .31 -0.22 -0.09 102 .50 -0.01 -021 .06 103 .61 .27 -0.21 -0.13 106 .50 -0.06 -0.12 -0.04 107 .51 -0.04 -0.08 . 11 109 .54 -0.03 -0.11 -0.01 110 .50 .00 -0.08 .08 112 .61 .51 -0.20 -0.07 113 .58 .62 -0.13 -0.07 115 .52 .01 -0.00 .04 120 .52 -0.03 -Q.10 .28 Station 22 121 .44 -0.07 .66 -0.15 122 .52 .34 .47 -0.20 123 .46 -0.04 .55 -0.12 126 . 50 -0.09 33 -0. 20 127 .46 -0.09 .77 -0.22 128 .45 -0.08 .80 -0. 21 129 .44 -0.06 79 -0.18 130 .47 -.0.10 .81 -0.24 131 .46 -0.05 .53 -0.13 132 .47 -0.11 .75 -0.22 133 .44 -0.06 .57 0. 13 135 .45 -0.08 .70 -0.18 136 .46 -0.09 .62 -0.15 137 44 -0.09 .60 -0.14 138 .48 -0.12 .56 -0.05 140 .43 -0.11 .54 -0.13 87

Table 17.Continued.

Station Sample Nwnbe' I II

23 142 .55 .51 .02 -0.12 143 .56 .69 -0.14 -0.12 144 . 57 -0. 16 -0. 10 145 52 .46 -0. 10 -0.07 146 .55 .41 -0.16 .03 147 .52 -0.01 -0.04 -0. 13 148 .59 .33 -0.21 -0.08 149 .53 -0. 11 -0. 12 -o. 16 150 .52 -0.07 -0.11 -0.11 151 .53 -0.04 -0. 16 -0.09 154 .54 39 -0.15 -0.01 155 .55 .44 -0.13 -0.10 156 .51 -0.16 -0.13 -0.06 159 52 -0.07 -0.08 -0.03 160 . 51 -0. 16 -0. 13 -0.02 88 grabs at station 6 show a high loading to any factor and this is to fac- tor 1.Station 8 shows almost no loading with two of the 20 grabs loading on factor numberl, The lack of significant response for these two stations is again thought to be based on the low numbers of individuals and low numbers of species. The agreement between R mode and Q mode is quite good. R mode extracted four species groups, two of which were found in the silty-sand, sandy-silt stations, one from the deep glauconite sand, station 10, and one from the beach sand, station 22. Q mode shows that by aligning the stations by species variation, the silty- sand stations show heavy loadings to two factors, factor 1 and fac- tor 2.Where the beach sand environment, station 22, is the only station to show high loadings to factor 3, the glauconitic sand environ- ment, station 10, shows high loadings to factor 4 and stands by itself. These three species groups represented by the nearshore beach sand, the intermediate depth silty-sand, and the deep glauconitic sand, represent three shelf communities. While the two sand sta- tions on the extremes of the depth range are consistent and distinct, the silty-sand, sandy-silt stations show intergradationwith two sub- groups being distinguishable; those in predominately silt area or silt- covered areas represented by stations 2,6, and 8 and those in silty- sand areas represented by stations 7,15, and 23.The- silt stations 6 and 8 do not align with any of the factors or species groups.The interrelationships of the seasonal stations indicated by the AIDN and 89 factor analyses are shown in Figure 22.

Regression Analysis

The 21 most abundant shelled molluscs in the samples were examined using regression analysis (Table 18).The nine environ- mental variables chosen for regression were depth; salinity, temper- ature, and dissolved oxygen of bottom water; calcium carbonate, organic carbon, sorting coefficient, percent silt, and percent sand of sediments. The remaining variables were highly correlated with one or more of the variables used and thus their use would have been redundant. Ten of the 21 species showed significant regression with one or more variables based on the F statistic; in only seven of these were the results meaningful. The relatively high percent variation explained by salinity for Olivella baetica, Acteocina eximia, and Cylichna attonsa is not biologically meaningful since salinity varies little throughout all the samples. Four of the remaining seven species show a relatively low response to depth.Depth variation accounts for 16 to 39 percent of the variation in abundance.This is generally unimpressive,Species such as Tellina salmonea and Odontogena borealis are abundant species found only at station 22 and station 10 respectively.These stations are both the shallowest and deepest sta- tions and are distinctive in regard to the sediment types, station 22 90 SEA GRANT STATIONS

440 301

440 00'

Figure ZZ.Interrelationships of the eight seasonal stations.One nautical mile equals 1 . 85 kilometers; 1 fathom equals 1.83 meters. 91

Table 18.Regression analysis for the 21 most abundant molluscs on nine environmental parameters.

Species Variable Cuni. % Variation F plained Statistic F. 01

Tellina salmonea Sorting 13.49 2.49 8. 53 %Sand 19.21 1.06 8.86

Odontogena borealis So/oo 14.59 2. 22 9.07 % Sand 17.98 4.97 9.33

Acila castrensis Depth 34.76 15.98** 7.56 Oxygen 42.09 3.66 7.60

Huxleyia znnuta Depth 38.86 24. 16** 7.35 % Sand 40.86 1. 25 7. 37 Magacrenella columbiana Sorting 16.59 3.78 8. 18 Oxygen 18.74 0.05 8.28

Cardita ventricosa % Sand 17. 23 9. 78 7. 20 Depth 19. 60 1.36 7. 21

Odontorhinacyclia Depth 16.30 9. 73' 7. 17 % Silt 18.54 1.35 7.18 Axinopsida seriata Organic Carbon 8.34 5. 11 7. 11 T°C 12.07 2.32 7.12 Thyasira gouldii Organic Carbon 18.32 09** 7. 22 Oxygen 21.52 1. 79 7. 24 Compsomyax subdiaphana CaCO3 11.78 4.81 7.39 % Sand 14.42 1.07 7.42 Macoma carlottensis S 0/00 26.17 7.44 8.02 Sorting 39.73 4.49 8.10 Macoma elimata CaCO3 20.62 8.83** 744 S 0/00 22.37 0.74 7.47 Bittiwn minutum Depth 14.05 2. 12 9.07 Oxygen 21.35 1.11 6.93 Polinices pallidus Sorting 13.12 1.81 9.33 T°C 15.47 0.31 9.65 Mitrella gouldi Organic Carbon 4.43 4.22 6.93 Sorting 7.54 3.02 6.93 Olivella baetica S o/oo 42. 27 16. 14** 8. 28 Organic Carbon48.48 0.40 8.40 Acteocinaexinla S 0/00 44.55 11.25** 8.86 Organic Carbon46. 20 0.40 9.07 92

Table 18.Continued.

Species Variable Cum. % Variation F Fxplained Variable F. 01

Cylichna attonsa S of oo 99.57 373577** 8.53 T°C 99.86 0.31 8.68 Dentalium rectius Depth 21.60 23. 14** 6.95 CaCO3 31.63 2.17* 6.95 Dentallum pretiosum Depth 14.30 1.83 9. 65 Salinity 33.48 2.88 10.04 Cadulus stearnsii Sorting 38.84 19. 05 7.56 39.19 0.17 7.60

*5% 93 being almost 100 percent beach sand while station 10 is 84 percent glauconite.Neither of these species shows a significant correlation to percent sand or to depth.Glauconite is suspected to be of bio- genic origin, formed in a reduction environment in foraminiferan testes.The size of the particles is in the same range as beach sand and the sediment particle size analysis used makes nb distinction be- tween the two.The treatment of these two sediments as the same thing in the regression analysis may in part account for the lack of significance for the sand species. 94

IV. DISCUSSION

The lack of seasonal changes in the abundance of infaunain the samples analyzed in this investigation is not surprising when the shelf environment is taken into consideration.The summer upwelling per- iod and the associated high production does not appear to have a short- term, large-scale effect on the infauna of the continental shelf.There is no seasonal pattern in the sediment organic nitrogen, organic car- bon or in the total carbon on the sea bottom (see Appendix I).The lack of seasonality in these and in the six sedimentary textural param- eters indicates that the production and the surface waters have long- term, more widespread effect on the bottom.The effect of upwelling on production during the summer period is possibly counterbalanced by the increased runoff and flow of materials from the rivers during the winter and spring.Preliminary indications of high velocity, bottom currents (Harlett, personal communication) may also be a major factor in the lack of a discernible seasonal pattern. The lack of seasonality in the fauna as a whole does not exclude the possibility of large-scale changes in particular species on the taxonomic groups not studied in this investigation.Thorson (1946, 1950) demonstrated the seasonality of benthic invertebrates and showed that the effect of the seasonality could be stretched over a long-term period because of differences in lengths of larval life. 95 Among the 35 major species of molluscs that were subjected to factor analysis, no seasonal pattern of growth of individuals or recruitment could be seen.The greatmajority of individuals of a given species were approximately the same size, and young were rarely found0 This was particularly striking in Acila castrensis and Tellina salmonea, the two most abundant species in the beach sand com munity at station 22,No young of either species was found over the course of the study; although adults were found in densities of ten to 350 per square meter in over 90 percent of the grabs.If these par- ticular species are long lived, and recruitments are scattered over a year or every other year, it is possible that recruitment could have been missed completely. Because measurements were not taken on all specimens, it is impossible to make statements about seasonal variation of individual species0 Longhurst's review (1964) of the state of benthic synecology discussed the difficulty of making comparisons between various benthic investigations.This difficulty is primarily due to the made- quacy of sample size in many previous stu.diesand to the variation in gear used and methods of analyzing the fauna. Previous studies on the west coast of North America such as those of Shelford etal. (1935), Hartman (1955) and G. L. Jones (1964) were largely semiquan- titative and the former two nonseasonal.Lievs study in Puget Sound (1968) was both quantitative and seasonal,Based on ten 0, 1m2 grabs 96 per season, Lie found that there was noseasonal variation in either the number of specimens per species or in the number of species. In addition, he found that there was no seasonality in thetotal bio- mass of the infauna although there wasconsiderable variation in individual weights of each sample. The presence of largewidely scattered individuals such as the echinoid Brisaster latifronscaused this.The same result had previously been demonstrated byBlegvad (1926), Steven (1930) and N, S. Jones (1952). For a comparison of species composition, density,and total number of species, the study of Steven (1930) is inappropriatebe- cause of the paucity of molluscan fauna,Stephen (1933) determined the distribution of the major molluscan species in theNorth Sea based on a thousand widely scattered samples. For his offshore zone, 40 to 60 meters in depth, pelecypoddensities range from three to 63 permeter2, scaphopods from 0 to 11 per meter2, and gastro- pods from 0 to 15 permeter2.Stephen's coastal zone, however, from just subtidal to 36 meters in depth, has bivalvedensities of 19 to 635 permeter2.The numbers have little meaning, however, to the present study because of the differentconditions that prevail in the two areas, the different depth rangestudied, and lack of sediment data.It does, however, give a general range of previously reported molluscan densitiesfrom a quantita- tive study. 97 The study of Holme (1953) in the English Channel. reports densi- ties of bivalves up to 206 permeter2,This is less than that reported by Ford (1923) in the same area where he found 576 permeter2.Part of the difficulty may be overestimate due to inadequate samples in the earlier study of Ford. Wigley and McIntyre (1964) in their study on the continental shelf (40 to 366 meters; 9 stations) south ofMartha's Vineyard, Massachusetts found to 525 individuals permeter2 for the molluscs and up to 5, 515 permeter2 witha mean of 2,418 per meter, for the total macrofauria. Gilbert F. Jones (1964) using 355.0 25m2 grab samples found that the molluscs averaged 16.5% of the total fauna by species and 12% by numbers on the continental shelf (90 meters) between Point Conception and San Diego, California,He found that north of Santa Barbar the bivalve Cardita ventricosa comprised about half of the wet weight standing stock (52.4 grams permeter2of a total standing stock of 103, 2 grams permeter2). Themean number permeter2 was 144,There was no apparent seasonal variation but replicate samples were not taken and in-station variationwas very high. Direct comparison of the data is impossible because of the slight overlap of the depth range (15 meters). Carey (in press) found in the quantitative study using an anchor- box dredge that at 75 meters the number of individuals ranged from just over 100 to over 4,000 permeter2 witha mean of about 1, 100 98 permeter2.The difficulty of using the anchor-box dredge as a quan- titative tool as shown by large in-station variation is compounded by the heterogeneity of sediment distribution.At 125 meters Carey also reported about 1, 100 individuals permeter2.The numbers at 150 meters off Newport ranged from approximately 200 to about 1800 per meter2.The edge of the shelf showed a considerable increase with numbers ranging from 500 to almost 5, 000 permeter2,with the mean number permeter2for eight stations ranging from 665 to 1,943, These values are higher than those found in this study which ranged from 247 to 988 with a mean of 597 permeter2.This is because Carey used a 0. 42 mm sieve size where a 1.0mm sieve was used in this study.Both values are lower than those found by Wigley and McIntyre (1964), and Sanders (1956) and Sanders etaL (1965) in the North Atlantic Ocean. Carey found that the infauna increased in both biomass and numbers seaward.Theopposite was found in this study with the maximum coming at 75 meters and decreasing seaward to 450 meters. Numbers of individuals reached a maximum in the silty-sand stations at 100 meters and decreased seaward.The dis- crepancies are possibly due to differences in the sampling gear and sieve size used in the two studies. Comparison of benthic biomass data with that of other areas is difficult since there are no standardized methods of measuring bio- mass. Alcohol, formaldehyde, and freshwet weights are used as 99 well as dry weights and ash-free dry weight.Wigley and Mcintyre (1964) found that the total wet preserved weight of the macrofauna off Martha's Vineyard, Massachusetts ranged between 24.3 and 1,355.7 gramsper:meter2with an average for eight stations of 266.3 grams permeter2.in this study wet weight ranged from 7.4 to 101.8 grams permeter2with an average of 36. 5,This is below that found by Wigley and Mcintyre on the East Coast. They found the weight of the macrofauna to be lowest at the deeper stations and highest at the shal- low ones.This is true of wet weight and ash free dry weight in the present study also. Lie (1969) in a study off the Washington coast found the ash free dry weight at 37 stations between 12 and 329 meters varied from 0.47 to 6.67 grams permeter2with a mean of 1.29 grams permeter2. At 70 percent of his stations the standing crop was less than 2. 0 grams permeter2.This is considerably less than that reported by Sanders (1956) in Long Island Sound, Barnard and Hartmn (1959) off Southern California, and G. F. Jones (1964) off Southern Cali- fornia. Shevdsov (1964), cited in Lie (1969), reported an edible

fraction of eight grams permeter2in the Gulf of Alaska.Lie ( 1969) states, however, that four stations greatly influenced the mean value given by Shevdsov. The exclusion of these stations gives a biomass of 2.7 grams per meter2,Carey (in press) reports a biomass on the nearshore shelf of 3.9 grams permeter2and a biomass of 4.5 grams 1 00 per:meter2at the shelf edge.In this study the ash free dry weight varied from 0, 9 to 4. 6 grams permeter2.The highest value came from the nearshore sand station and the lowest value from the deep sand station, stationlO,The mean weight was 2.57 grams per me- ter2.Therefore the shelf of the Pacific Northwest seems to be low in biomass when compared to the Atlantic at the same latitude.It seems strange that in an area with high surface production such as the coastal waters of Oregon the standing stock and number of benthic individuals permeter2is so low,This would be true if the high surface production never reached the bottom due to currents which carried it off the shelf. The only work on benthic infaunal communities off the Northwest coast is that of Lie and Kelly (1970),Using factor analysis, an index of affinity, and Fager'srecurrent group analysis (Fa-ger, 1957; Fager and Longhurst, 1968) they delineated three communities off the Wash- ington coast.Lie and Kelly characterize three communities lying parallel to the coast of Washington; a shallow sand community from 15 to 83 meters depth, a muddy sand community from 50 to 164 me- ters depth, and a deep-water mud community from 117 to 317 meters depth.These three communities have their analogues ofL the coast of Oregon, but there are some interesting differences.Station 22 of the present investigation would be analogous to the near shore sand community described by Lie. 101

The most abundant bivalve in Lie's nearshore community (Un-. published data) was Tellina salmonea.This bivalve is second in abundance in the shallow sand community at station Z2; however, the next four species described in Lie's first community, Macoma expal sa, Tellina buttoni, Siliqua spatula, and Axinopsida sericata, do not occur at station 22,In addition, the most abundant species at station 22 is Acila castrensis.This species is three times as abundant in terms of numbers as Tellina salmonea, The intermediate community has as its five dominant bivalves Yoldia ensifera, Axinopsida sericata, Macoma elimata Acila castrensis, Nucula belloti, and Compsomyax subdiaphana. These six species are present to some degree at sta- tions 7,15, and 23.The order of abundance, however, is different, Compsomyax subdiaphana plays a major role at the three stations while Acila castrensis is present in very low numbers. At stations 7 and 23 Macoma carlottensis is the much more abundant species while Macoma elimata is most abundant at 15,These stations are located within tbe depth range described by Lie of 50tol64 meters and in spite of this one, specific difference can be considered as part of his muddy-sand community. The deep mud station described by Lie has as its five dominant bivalves, in order of abundance, Axinopsida sericata, Adontorhina crclia, Macomacarlottensis, Thyasira gouldii, and Tellina carpen ten.At stations 2,6, and 8, which lie within the depth range 1 02 described by Lie and Kelly and are mud stations, the first four spe-. cies are present in abundancewhile the Lie's dominant species, Axinopsida sericata and Adonthorhina cyclia are co-dominants in both station 2 and station 6.At station 8, which has a very low mol- luscan fauna, Thyasira gouldii and Axinopsida sericata are also the co -dominants. Therefore, Lie and Kelly's described communities stand up quite well on the Oregon coast with exceptionsf certain species changing abundance in certain communities and one, Acila castrensis, becoming dominant in a community in which it does notexist in Wash- ington.To these three communities a fourth community is added. This community is the glauconitic sand community at 450 meters. By difference index, similarity index, and factor analysis both in Q and R. modes, this community has been shown to be distinct from those already described,The glauconitic sand-siltcommunity has as its three dominant bivalve members Crenella discussata, Odon- togena borealis, and Huxleyia minuta. In the sense that these four communities lie in bands more or less parallel to the coastline the distribution of species on the Oregon shelf can be considered as a continuum as defined by Curtis and Mclntosh(1951) and Whittaker (1967, 1970), The ecotones between these communities may be quite broad and difficult to distinguish (the objectionraised by Lindroth (1935) 1 03 and Stephen (1933, 1934)), but this does not eliminate these communi- ties from consideration.The distinctiveness and broad distribution of the communities is evidenced by their presence on both the Oregon and Washington coasts.The species, however, are indeed distributed along an environmental gradient, that of depth and of particle size going from coarse to very fine sediments and in some places to coarse again.Within the limits of their distribution along physical parameters, the distribution of particular species may be based on biotic factors.The species are more highly correlated with each other than with any environmental parameters. The results of factor analysis also indicatethat a large portion of the distribution is unaccounted for by the distribution of the major community dominance considered in the analysis.Predators or prey may be more highly correlated with the species in question than are those species with high factor loadings.Species with low numbers that were not considered in the analysis may be the major reason for particular distrib.i.tional patterns.Lie and Kelly (l97Q) found that 58. 7 percent of the total variation was accounted for in R mode fac- tor analysis by four factors.This was considerably less than was found by Colebrook (1964) for four factors in a zooplankton population. The total variance in this study that was accounted for by four factors in R mode was 71.3 percent.It appears that the benthic communities 104 on the Oregon shelf are complex and may have a very high degree of species interaction. 105

V. CONCLUSIONS

There are four distinct communities on the Central Oregon shelf lying in four distinct sediment types more or less parallel to the coastline.These communities can be found in the distinct substrates of beach sand, mixed silt and sand, silt, and glauconitic sand.The fauna, when compared with the East Atlantic shelf and the shelf off Southern California, is low in number of species, number of individ- uals, and in biomass in spite of the high production in the surface wa- ters over the shelf.There is no significant change in the faunal com- position in numbers of biomass with season at any of the stations from 75 to 450 meters. The sediment environmental parameters such as organic carbon, organic nitrogen, and particle size show no signifi- cant seasonal change. There appears to be a high degree of species interaction that indicates complex benthic communities. The broad findings of this investigation have use in a number of ways. The apparent lack of seasonality in the infauna and thelow standing crop have broad implications for fisheries biology.The understanding of the distribution of the four communities on the Oregon shelf may be of help in understanding the distributional pat- terns of migratory demersal fishes.The identification and enumera- tion of the infauna forms a base for future work on fish stomach an- alysis to determine selective feedings by season, sex, or size. 106 Lastly, the methodology and analysis of the data shows some of the things that can be done with adequate quantitative data and points to the need for additional methods to analyze particular species in refer ence to their total environment including the biotic and abiotic 107

BIBLIOGRAPHY

Barnard, J, L. and 0. Hartman,1959,The sea bottom off Santa Barbara, CaliLornia: Biomass and community structure0 Pa- cific Naturalist 1(6):l-16,

Barnard, J. L. and G. Jones,1959,Benthic ecology of the mainland shelf of southern California.State of California Water Pollution Control Board Publication 20:265-429.

Bartlett, M. S.1947.The use of transformations.Biometrics 3: 39-52.

Blegvad, H.1925.Continued studies on the quantity of fish food in the sea bottom,Report Danish Biological Station 31:27-56.

Byrne, J. V. and D. A. Panshin.1968.Continental shelf sediments off Oregon. Sea Grant Publication Cooperative Extension Service Oregon State University.4 p. Carey, A. G., Jr.1971,Ecological observations on the benthic invertebrates from the Central Oregon continental shelf.In: Bioenvironmental studies of the Columbia River Estuary and adjacent ocean ed. by D. L, Alverson and J. N. Wolfe, Atomic Energy Commission.(In press)

1971.Techniques for sampling benthic organisms. In:Bioenvironmental studies of the Columbia River Estuary and adjacent ocean ed. by D. L. Alverson and J. N, Wolfe. Atomic Energy Commission,(In press) Carey, A. G., Jr. and R. R. Paul,1968. A modification of the Smith-McIntyre grab for simultaneous collection of sediment and bottom water.Limnology and Oceanography 13(3):545- 549. Carey, A. G., Jr. and D, R. Hancock.1965. An anchor-box dredge for deep-sea sampling.Deep -Sea Research 12(6):983 -984,

Colebrook, J. M.1964.Continious plankton records: a principal component analysis of geographical distribution of zooplankton. Bulletin of Marine Ecology 6:78-100, 108

Curl, H.1963.Analysis of carbon in marine plankton organisms. Journal of Marine Research 20(3):181-188, Curtis, J. T. and R. P. McIntosh,1951, An upland forest continu- um in the prairie-forest border region of Wisconsin, Ecology 32:467-469. Davis, F. M.1923.Quantitative studies on the fauna of the sea bot- tom Part 1.Preliminary investigations of the Dogger Bank. Fisheries Investigations London 2(6).

1925,Quantitative studies on the fauna of the sea bottom,2: Results of investigations in the southern North Sea 1921-1924.Fisheries Investigations London 2(8).

Draper, N. R. arid H. Smith.1966,Applied regression analysis. John Wiley and SonsN. Y.407 p.

Emery, K. 0,1938,Rapid method of mechanical analysis of sands, JournalSedimentology Petrology 8:105-111,

Fager, E. W,1957,Determination and analysis of recurrent groups, Ecology 38(4):586-595.

Fager, E. W. and A, R. Longhurst.1968.Recurrent group analy- sis of species assemblages of demersal fish in the Gulf of Guinea,Journal of Fisheries Research Board of Canada 25: 1405-1421,

Ford, E.1923,Animal communities of the level sea-bottom in the waters adjacent to Plymouth.Journal of the Marine Biological Association of the United Kingdom 13:164-224,

Fowler, G. A. and L, D. Kuim.1966.1966, A multiple corer. Limnology and Oceanography 11(4):630-633. Fjarlie, R. L, I.1953, A seawater sampling bottle.Journal of Marine Research 12:21-30,

Gallardo, U. A.1965,Observations on the bitin profiles of three 0. 1 m2 bottom samplers.Ophelia 2(2):319-322.

Harlett, J.1970,Graduate Student,Department of Oceanography Oregon State University.Corvallis.Personal communication, 109

Hartman, 0.1955.Quantitative survey of the benthos of San Pedro Basin Southern California.Part I.Preliminary Results Allan Hancock Pacific Expeditions 19:1-185,

Hedgpeth,J. W.1957,Sandy beaches.In:J. W. Hedgpeth ed,, Treatise on Marine Ecology and Paleoecology, Geological So- ciety of America, Memoirs 67(l):587-608.

Holme, N. A.1953,The biomass of the bottom fauna in the English Channel off Plymouth,Journal of the Marine Biological Asso- ciation of the United Kingdom 32:1-50. 1959, A hopper for use when sieving bottom sam- ples at sea,Journal of the Marine Biological Association of the United Kingdom 38 (3):525-528.

Imbrie, J. and T. H. van Andel.1964,Vector analysis of heavy- mineral data.Geological Society of America Bulletin 75:113 1- 1156.

Jones, G. F.1964,The distribution and abundance of subtidal benthicMollusca on the Mainland shelf of Southern California. Malacologia 2(l):43-68. Jones, M. L.1961. A quantitative evaluation of the benthic fauna off Point Richmond California,University of California Pub- lications in Zoology 67(3):219-320.

Jones, N, 5,1952,The bottom fauna and the food of flat fish off the Cumberland coast,Journal Animal Ecology 21:182-205,

1956,The fauna and biomass of a muddy sand de- posit off Port Erin Isle of Man.Journal of Animal Ecology 25:217-252.

Kaiser, H. F.1958,The varimax criterion for analytic rotation in factor analysis.Psychometrika 23:187-200,

Krumbein, W. C. and F. J. Pettijohn.1938,Manual of Sedimentary Petrography. Appleton-Century-Crofts Inc.N. Y.549 pp. 265 figures.53 tables,

Levins, R. 196B. Evolution in changing environments. Princeton University Press.120 pp. 110

Lie, U.1968. A quantitative study of benthicinfaunain Puget Sound WashingtonU. S. A. in 1963-1964.Fiskeridirektoratets Skrifter Serie Havunderske1ser 14:229-556.

1969,Standing crop of benthic infaunain Puget Sound and off the coast of Washington.Journal of the Fisher- ies Research Board of Canada 26:55-62. Lie, U. and J. C. Kelly. 1970.Benthic Infauna Communities off the Coast of Washington and in PugetSound: Identification and Distribution of the Communities.Journal of the Fisheries Re- search Board of Canada 27(4):62l-651.

Lindroth, A.1935.Die Associationen der marinenWeichboden. Zoologiska bidrog fran Uppsala 15:331-336.

Longhurst, A. R.1957,Density of marine benthic communitiesoff West Africa.Nature 179:542-543.

1959.The sampling problem in benthicecology. Proceedings New Zealand Ecological Society6:8-12. 1964. A review of the present situationin benthic synecology.Bulletin Institute Oceanography Monaco63:1-54.

MacArthur, R. H.1965.Patterns of species diversity.Biological Reviews 40:510-533, Mclntire, D. C. and W. S. Overton.1970.Distribution patterns in assemblages of attached diatoms from YaquinaEstuary, Oregon. Submitted to Ecology, 3 1 p. Maloney, N. J.1965. Geology of the continental terraceoff the central Oregon coast.Ph. D. Thesis Departmentof Oceanog- raphy Oregon State UniversityCorvallis,233 p.

Overton, 5.1969.Professor. OregonState University.Depart- ment of Statistics. AIDN experimental program.Corvallis, Oregon. Packard, E. L.1918. A quantitative analysis ofthe molluscan fauna of San Francisco Bay.University of California Publicationsin Zoology 18:299-336. 111

Petersen, C. G. J.1913.Valuation of the sea II.The animal communities of the sea bottom and their importance for marine zoogeography.Report Danish Biological Station Zl:44 p.

1918.The sea bottom and its productionof the fish- food. A survey of the work done in connection with the evalua- tion of the Danish waters from 1883-1917,Report Danish Bio- logical Station 25:62 pp. 1924. A brief survey of the animal communities in Danish waters. American Journal of Science Series 57(41): 343-35 4,

Petersen, C. G. J. and P. B. Jensen.1911,Valuation of the sea. I.Animal life of the bottom, its food and quantity.Report Danish Biological Station 20:81 pp. Reish, D. J.1959, A discussion of the importance of the screen size in washing quantitative marine bottom samples. Ecology 40: 307- 309.

R.unge, E. J.1965.Continental Shelf Sediments.Columbia River to Cape Blanco, Oregon.Ph. D. Thesis,Department of Oceanog- raphy. OregonState University,143 p.

Sanders, H, L. 1956, Oceanography of Long Island Sound.19 52- 1954,X.Biology ofmarine bottom communities,Bulletin of Bingham Oceanographic Institute 15:345 -414. SandersH. L,, R. R. Hessler, andG. R. Hampson.1965. An introduction to the study of the Gay Head-Bermuda transect, Deep-Sea Research 12:845-867. Shelford, V. E. and E. D. Towler,1925. Animal communities of the San Juan Channel and adjacent areas.Publications Puget Sound Marine Biological Station 5:33-73, Shelford, V. E., A. 0. Weese, L, A. Rice, D. I. Rasmussen, A. MacLean, N. Wismer, and J. H. Swanson.1935. Some marine biotic communities of the Pacific coast of North America.Ecological Monographs 5:25 1-3540

Shepard, F. P.1963,Submarine Geology.2nd Edition,Harper and Row, New York.557 p. 112

Shevtsov, V. V.1964, 0 kolichestvennorm raspredelenii donnoi fauny v zalive Alyaska.Trudy vses, nauchno-issle. Inst. morsk. ryb. Khoz. Okeanogr. 49:l07l12. Simpson, E. H.1949, Measurement of diversity.Nature 163:688. Smith, W. and A. D. McIntyre.1954, A springloaded bottom sampler.Journal of the Marine Biological Associationof the United Kingdom 33(l):257-264.

Snedecor, G. W. and W. G. Cochran,1968.Statistical Methods. 6th Edition.Iowa State University Press. Ames, Iowa. 593 pp.

Stander, J. M.1970.Diversity and Similarity of Benthic Fauna off Oregon. M. S. Thesis, Department of Oceanography, Oregon State University, Corvallis, Oregon.

Stephen, A. C.1933,Studies on the Scottish Marine fauna: The natural faunistic divisions of the North Sea asshowri by the quantitative distribution of the mollusca.Transactions Royal Society of Edinburgh 57:601-616.

1934,Studies on the Scottish marine fauna: Quan.- titative distribution of the echinoderms and the natural faunistic divisions of the North Sea.Transactions Royal Society of Edinburgh 57:777-787.

Steven, G. A,1930,Bottom fauna and the food of fishes,Journal of the Marine Biological Association of the United Kingdom l6:677706.

Thamdrup, H. M.1938.Der van Veen-Bodengreifer.Vergleich- sversuche uber die Leistungsfahegkeit des van Veen-und des Petersen Bodengr eifer s.Journal Conseil International Ex- plorationMer 13:Z06-21Z,

Thorson, G.1933,Investigations on shallow water animal com- munities in the Franz Joseph Fjord (East Greenland) and ad- jacent waters: Meddelesser om Grnland 100(2):l-68. 113

Thor son, G.1934,Contributions on the animal ecology of the Scoresby Sound fjord complex (East Greenland): Meddelesser omGrnland l00(3)l67. 1946.Reproduction and larval development in Danish marine bottom invertebrates, Meddeles ser Kom- missionen Danmark, Fiskeridirektoratets Sksifter Havunder skelse, Serie Plankton 4:1-523. 1950. Reproductive and larva.1 ecology of marine bottom invertebrates,Biological Reviews 25:1-45, 1951, Animal communities of the level sea bottom, Annales Biologie 27(7):48l-489.

1957,Bottom communities,In:J, W. Hedgpeth ed.Treatise on Marine Ecology and Paleoecology, Geological Society of America Memoirs 67(1):46l-534. Whittaker, R. H, 1967.Gradient analysis of vegetation,Biological Reviews 42:207-264,

1970.Communities and ecosystems.Collier- Macmillan Co.Ltd,, London,162 pp.35 figures,6 tables,

Wigley, R. L.1967,Comparative efficiencies of the van Veen and Smith-McIntyre grab samples as revealed by motion pictures, Ecology 48:168-169. Wigley, R. L. and A. D. McIntyre.1974. Some quantitative com- parisons of offshore meiobenthos and macrobenthos south of Martha's Vineyard.Limnology and Oceanography 9(4):485- 493,

Zenkevitch, L., V. Brotzky, and M, Idelson,1928,Materials for the study of the productivity of the sea-bottom in the White, Barents and Kara Seas.Journal Conseil International Explora- tion Mer 3:371-379, APPENDICES APPENDIX I SEA GRANT ENVIRONMENTAL DATA

Environmental Cruise Stations prranieter 2 6 7 8 10 15 22 23

Depth (meters) 6810 190 145 100 197 494 104 75 104 6901 200 146 99* 200 446* 100 73 100* 6904 189 150 100 200 443 102 75 102 6907 183 150 99 183 452 100 73 102

Salinity %o) 6810 33.89 33.80 32.82 33.82 33.92 33.74 33.48 33.69 6901 33.87 33.52 32.87* 33.82 339Ø* 33.38 32.68 33.32 6904 33.96 34.01 33.78 34.02 33.87 33.74 33.16 33.74 6907 34.04 33.04 33.99 34.03 34.13 34.01 33.99 33.97

Temperature (°C) 6810 7.66 8.01 8.20 8. 26 5.66 8 18 8. 67 8.24 6901 8.19 8.20 12.04 12.04 5.61 8.78 9.21 8.88 6904 6. 60 6. 15 7.43 6.58 5.51 6.87 7. 50 7.04 6907 6. 11 6. 38 7.04 6.47 5.52 7.02 6. 70 7. 18

Oxygen (nil/i) 6810 2.49 2.31 3. 15 2.54 2.39 2.13 3.87 3.12 6901 2.98 4.13 5.61* 3.08 1.23* 4.08 5.59 4.18* 6904 2.34 2. 10 1.83 2. 08 2.98 3. 63 3.94 2.60 6907 2.13k 1.65 1.96 1.90 0.87 1.57 1.57 2.23

Silicates (1m/i) 6911 29 39 40 19 57 3 34 34 7002 21 27 15 38 32 17 10 21 6904 41 47 47 48 33 41 31 40 6907 54 54 56 58 51 63 56 50 APPENDIX I (Continued)

Environmental Cruise Stations parameter 2 6 7 8 10 15 22 23

Phosphate(1j.m/l) 6911 1.91 2.33 3.05 2.16 2.70 1.98 1.89 3.09 7002 2.33 3.03 2.03 3.92 1.70 2. 18 1. 83 2.31 6904 2.89 3. 24 3.03 3. 10 2. 75 2. 79 2. 26 2. 89 6808 2.56 2. 69 2. 63 2.58 2.02 2. 60 2.43 2. 60

Nitrates(IJm/l) 6911 21.2 28.4 26.6 11.5 33.7 23.7 25.6 24.2 7002 14.8 17. 8 9.4 24.5 20.0 11. 2 6. 6 17.0 6904 28.67 31.70 28.10 33.35 27.33 26.08 14.41 26.49 6907 34.1 25.4° 33.1 33.4 32.2 34.0 31.2 31,0

% Total Carbon 6810 1.05 1.71 0.62 1.57 1.06 0.52 0.07 1.30 6901 1.08 1.86 0.51* 1.67 1.12* 0.37 0.11 1.30* 6904 1.21 1.48 0. 65 1.94 0. 75 0.44 0. 18 1.40 6907 0.81 1.89 0.76 1.56 0.49 0.46k 0.10 1.51

% Calcium 6810 .056 .056 .093 .098 .000 .002 .023 .042 Carbonate 6901 .047 .042 .032* .102 .010* .009 .000 .046* 6904 .052 .051 .046 .084 .014 .028 .028 .069 6907 .047 . 070 . 014 .074 .000 .049k .019 028

% Organic Carbon 6810 994 1. 654 0. 527 1.472 1. 06 0.518 0.047 1. 258 6901 1.033 1.82 0.578* 1.568 1.11 0.361 0.1101 1.254* 6904 1.158 1.43 0.604 1.856 0.736 0.421 0.152 1.331 6907 .763 1.82 0.744 1.486 0.49 0.411k 0.081 1.482

Sediment 6810 1034 1885 548 1555 927 468 8 1160 Organic Nitrogen 6901 1015 1675 565* 1720 498* 446 12* 1110 l.L gin Nitrogen 6904 1170 1390 715 1765 656 397 11 862 gin Sediment 6907 1210 1710 667 1455 537 497 12 922 APPENDIX I (Continued)

Environmental Crtise Stations parameter 2 6 7 8 10 15 22 23

Folk and Ward 6010 4.76 7. 14 4. 22 7.70 2. 66 3. 33 1. 66 6.48 Mean Particle Size 6901 4.37 7. 12 3,57* 7.06 3.05* 2. 60 2.04 6.50* 6904 4.85 6.68 5.09 7.49 1.92 3.01 1.88 5.38 6907 4.52k 6.45 3.91 6. 63 2.76k 7.01* 1.93 3.67 Folk and Ward 6810 3.05 2. 68 2. 68 3. 13 2. 63 2. 27 0.48 3. 29 Sorting Coefficient 6901 2. 17 2.73 1.48* 2.61 2. 20* 1. 27 0.44 2,52 6904 3.24 2.91 2.88 2.79 2.39 1,40 0.45 3.19 6907 2, 61 3. 11 2. 20 2. 06 1. 83 2.30 + 0.43 2. 62 Inman 6810 3.47 6.31 3.25 6.65 1.63 2.83 1.70 5.13 Median Grain Size 6901 3.47 6. 12 3. 10* 6. 26 2. 19* 2. 60 2. 06 5. 23* 6904 3.49 5. 37 3.95 6. 63 1. 59 2.74 1.89 4.51 6907 3.43k 5.68 3.24 6.25k 2.61k 2.95k 1.91 4.98 % Clay 6810 15.67 34.70 13,35 39.76 9.17 9.23 0.00 26.64 6901 12.75 33.47 6.50* 33.81 4.98* 5.58 2.07 28. 17* 6904 16.08 27,86 17,20 38.25 7.96 5.27 0.93 18.22 6907 14.46k 25.10 9.60 29.50k 8.96 9.04k 0.73 21.39 % Silt 6810 16.68 64.07 24.84 59.14 11.91 11.67 0.00 49.62 6901 16.76 65.09 21.24* 64.15 27.45* 7.45 0.53 5O99* 6904 18.92 64.53 21.50 60.46 6.40 10.66 0.35 44.62 6907 14.82k 72.62 20.92 64.69k 6..94 4.80k 0.51 47.30

% Sand 6810 67.65 1.22 61.80 1.14 78.91 79.10 100.00 23.75 6901 70.49 1.44 72.26* 2.04 67.60* 86.97 97.40 20.83 6904 64.99 7. 60 51.30 1. 29 84.07 84.07 98.72 20.84 6907 70.72 2.26 69.51 5.81k 84.09k 86.15 98.76 37.15

KEY 1 6808, 1° = 6901, 1' = 6911, 1* = 7002, 1= extrapolated 117

APPENDIX II LIST OF SPECIES

Phylum Mollusca Class Scaphopoda Family Dentaliidae Dentalium rectius Carpenter Dentaliimi pretiosuxn Sowerby Family Siphonodentalidae Cadulus stearnsii Pilsbry and Sharp Cadulus californicus Pilsbry and Sharp

Class Pelecypoda Order Protobranchia Family Nuculidae Acila castrensis (1-linds) Nucula tenuis Montagu Family Nuculanidae Nuculana austini Oldroyd Nucuiana rninuta (Fabricius) Yoldia ensifera Dali Yoldia thraciaeforrnis Storer Order Prionodontida Family Nucinellidae Huxleyia minuta Dali Order Pteroconchida Family Mytilidae Crenella discussata (Montagu) Megacrenella columbiana (Dali) Musculus nigra Gray Order Heterodontida Family Carditidae Cardita ventricosa Gould Family Kellidae Odontogena borealis Cowen Family Montacutidae Pseudopythina 'A" Family Lucinidae Lucinoma anmilata (Reeve) Family Thyasiridae Adontorhina cyclia Berry Axinopsida serricata (Carpenter) Thyasira bisecta Conrad Thyasira gouldii (Philippi) Family Cardiidea Nemocrdiuin richardsoni (Wliiteaves) Family Veneridae Compsomyax subdiaphana Carpenter Psephidia lordi Baird 118

Family Tellinidae Macoma carlottensis Whiteaves Macoma elimata Dunnill & Coan Tellina carpenteri Dall Tellina salmonea Carpenter Order Eudesmodontida Family Pandoridae Pandora filosa Carpenter Family Lyonsiidae Lyonsia pugettensis Dali Family Thraciidae Thracja trapezoides Conrad Order Septibranchia Family Cuspidariidae Cardiomy pectinata (Carpenter) Cardiomya planetica (Dali)

Class Gastropoda Order Diotocardia Family Trochidae Solariella varicosa (Mighels and Adams) Solariella nuda Dali Order Monotocardia Family Eulimidae Balcis sp. Family Epitoniidae Epitoniurn caamanoi Dali and Bartsch Epitonium tinctum (Carpenter) Epitonhun acrostephanw Dall Family Turritellidae Tachyryncus lacteolus Carpenter Family Cerithiidae Bittium minutum (Carpenter) Family Naticidae Poljices pallidus (Broderip and Sowerby) Sub Order Neogastropodia or Stenoglassa Family Muricidae Boreotrophon dalli Kobelt Family Neptuneidae E.xilioidea rectirostris Carpenter Mohada Dali Neptunea liratus (Gmelin) Family Columbellidae Mitrellaouldi Carpenter Family Nassariidae Nassarius fossatus (Gould) Nassarius mendicus (Gould) Family Olividae Olivella baetica Carpenter Olivella biplicata (Sowerby) 119

Sub Order Toxoglossa Family Turridae Oenopota sp. Ophiodermella rhines Dali ppiodermel1a incisa Carpenter Rectiplanes thaiaea Dali Mangelia sp. Order Tectibranchiata Family Pyramidellidae Odostomia sp. Turbonilla pedroana Dali and Bartsch Twbonilla aurantia Carpenter Family Scaphandridae Acteoclna culcitella (Gould) Acteocina exirnia Baird Cylichna attonsa Carpenter Family Acteonidae Acteon punctocaeiatus Carpenter

Phyllum Arthropoda Class Crastacea Order Cumacea Family Lampropidae Hemilamprops californiensis Zimmer Diastylis jellucida Hart Diastylis paraspinulosa Zimnier Diastylis bideutata Caiman Diastylis dalli Diastylis "A' Family Colurostyiidae Colurostylis occidentalis Calman Family Leuconidae Eudorella pacifica Hart Lucon longirostris Family Nannastacidae Campylaspis rubicunda (Lillieborg)

Phyllum Echinodermata Class Ophiuroidea Order Ophiurida Family Ophiuridae Ophiura lukeni (Lyman) Qphiura sarsiLutken Ophiura sp. Family Amphiuridae UniopluspJs (Clark) Unioplus euryaspis (Clark) Axipgnathus pugetana (Lyman) Amphiuridae sp. APPFNDIX III STATION TO STATION AIDN PARAMETERS BASED ON MOLLUSCAN FAUNA ONLY SIMILARITY

Station 2 6 8 10 15 22 23

2 1.000 6 .790 1.000 7 .569 .513 1.000 8 .802 .740 .818 1.000 10 .002 .019 .008 .054 1.000 15 .308 .453 .558 .604 .141 1.000 22 .011 .080 .072 .042 .000 .173 1.000 23 .480 .383 .886 .621 .007 .382 .065 1.000 DIFFERENCE

2 1.000 6 1.216 1.000 7 1,320 1.313 1,000 8 1.221 1,225 1,325 1,000 10 1.944 1.93 1.91 1.859 1,000 15 1.427 1.383 1.169 1.374 1.765 1.000 22 1.918 1.783 1.778 1.867 1.991 1.712 1.000 23 1.374 1.413 1.109 1.443 1.922 1.284 1.805 1.000 WEIGHT PD MEAN NICHE BREADTH Station B2 B5 2 3. 8618 3. 3596 6 4. 5127 3,9339 7 4. 2324 3.8153 8 4. 0826 3. 5919 10 1.5242 1,3204 15 3. 7852 3.3166 22 1,8089 1.5774 23 4.0983 3,6489 APPENDIX IV SEASON TO SEASON AIDN PARAMETERS BASED ON MOLLtJSCAN FAUNA ONLY SIMILARITY Fall Winter Spring Summer Fall 1.000 Winter .880 1.000 Spring .893 .951 1.000 Summer .934 .852 .850 1.000

DIFFERENCF Fall 1.000 Winter 1.081 1.000 Spring 1.074 1.039 1.000 Summer 1.055 1.097 1,093 1.000

WEIGHT ED MEAN NICHE BREADTH Fall 3.546 3.417 Winter 3.679 3.496 Spring 3.791 3.653 Summer 3.556 3.438 Appendix 5.Number of Individuals per Species Species Code 1 2 3 4 5 6 7 8 9 10 11121314 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Station2 1 1 1 I, 5 15259 4 13 13 18 3 3 515 Station6 1 1 2 6 4290 2 1 1 6 2 2 Station7 5 2 8 1 11 5 19 6 2126 16 31 1650 3 .3 5 .1 1 Station8 1 1 1 4 7 7 1 1 1 Station 10 2 9 11089 2 82 1 2 Station 15 11 17 6 37 2210 29 6 1422 104 2 16 31213 4 A Station 22247 1 17 16 72 4 Station 23 10 216 1 25 70 2 14 7 3 3 53 8817 :3 5 1 1 TotalNo.273 6 221322 915618936201251715147211 414551011712679317213 316 182 21218 1 2 3637383940414243444546474849 505152535455565758596061626364656667686970 Station2 2 1 5 1 1 9 2 8 Station6 4 46 1 8 1 1 2 2 Station7 1 5122 2 70 93 51 1 32217112231 11 Station8 1 3 1 5 2 j. StationlO 1 1 2 1 8 1 Station 15 5 3 6 98 2 3 3 3 1 2 1 4 1 42 211 2 2 4 Station 22 1 45 1 87 13 5 2 3 1 1 Station 23 2 4 13 2 40 6 2 1 2 1 5 8 11 2 1 9 Total No. 2 1 1 14 28 142 3 2 1 307 111 99 13 43 4 1 29 107 117 242 394 18 1 2 1 36 3 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 Total Number Station 2 1 1 42 3 21 2 4 1 300 Station6 1 4 43 10 2 10 3 294 Station7 1 1 1 1205 22 65 12 3 16 713 Station8 1 5 10 1 4 1 5 1 2 67 Station 10 7 2 6 3 429 Station 15 1 1 1 18 27 14 4 4 21 579 Station 22 1 1 1 29 2 6 556 Station 23 1 1 1180 12 52 7 13 1 6 686 Total No. 2 7 5 2 2 2 1 1 1 3 498 29 61 10 165 13 34 332 57 3624 * No species for number 50 Appendix 6.Percent Frequency of Occurrence Species Code

1 2 3 4 5 6 7 8 9 tO 1112 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Station2 5 S 5 5 25 5 50 65 10 35 30 45 15 10 111 40 Station6 5 5 5 30 65 90 5 5 5 20 10 10 Station7 25 10 30 5 40 20 55 25 65 60 85 60 30 70 15 iS 20 5 5 Station8 5 5 5 5 15 25 20 5 5 5 Station 10 5 35 75 35 10 10 70 S 10 Station 15 15 70 5 90 55 40 65 25 30 50 85 10 45 10 40 45 20 5 Station 22 100 5 60 60 90 10 Station 23 20 10 55 5 10 15100 10 40 10 15 iS 85 30 40 15 20 5 5 %of Total Grabs204111914264149247833371313248152011118281914711 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50*51 52 53 54 55 56 57 58 59 60 6162 63 64 65 66 67 68 69 70 Station2 10 5 20 5 5 5 5 35 5 40 Station6 20 85 40 S 5 5 10 Station7 5 15 30 10 10 80 40 15 25 5 5 10 105 25 45 10 101 5 55 Station8 5 15 5 25 10 5 Station 10 5 5 10 5 30 S Station 15 15 5 25 100 10 5 15 105 10 5 5 5 20 10 10 40 20 10 5 20 Station 22 5 85 5 80 45 25 10 10 5 5 Station 23 5 15 40 10 80 20 10 5 10 5 25 25 40 5 5 45 %of Total Grabs 111698121158165821311264191111949111222 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90% of Species Present Station 2 5 5 70 10 65 10 20 5 37. 1 Station 6 5 20 65 40 10 40 15 30.3 Station7 5 5 5 95 60 85 50 15 55 56.2 Station8 5 15 40 5 205 255 10 27.0 Station 10 25 10 20 15 21. 3 Station 15 5 5 5 60 70 55 20 20 70 52. 8 Station 22 5 5 5 65 10 25 24. 7 Station 23 5 5 5 100 30 65 45 45 5 45 50.6 %of Total Grabs 12 4111111154 820 441 91718 129 * No species for Number 50